Skip to main content

Full text of "Harvey Sacks Lectures On Conversation, Volumes I, II The Estate Of Harvey Sacks ( 1995)"

See other formats



Harvey Sacks 



Gail Jefferson 

With an Introduction by 

Emanuel A. Schegloff 



Lectures on Conversation, Volume I, II Harvey Sacks 
© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6 

© 1992, 1995 by The Estate of Harvey Sacks 
Introduction © 1992, 1995 by Emanuel A. Schegloff 


350 Main Street, Malden, MA 02148-5020, USA 

9600 Garsington Road, Oxford OX4 2DQ, UK 

550 Swanston Street, Carlton, Victoria 3053, Australia 

All rights reserved. No part of this publication may be reproduced, stored in a 
retrieval system, or transmitted, in any form or by any means, electronic, 
mechanical, photocopying, recording or otherwise, except as permitted by the UK 
Copyright, Designs, and Patents Act 1988, without the prior permission of the 

First published as two volumes 1992 

First published as one paperback volume 1995 

5 2006 

Library of Congress Cataloging-in-Publication Data 

Harvey Sacks : lectures on conversation / edited by Gail Jefferson; with an 
introduction by Emanuel A. Schegloff. 
p. cm. 

Includes bibliographical references and index. 

ISBN 1-55786-705-4 (pb) 

1 . Conversation. 2. Oral communication. I. Jefferson, Gail. 

P95 .45 .H37 1992 

302.3' 46— dc20 91-45536 


ISBN-13: 978-1-55786-705-6 (pb) 

A catalogue record for this title is available from the British Library. 

Printed and bound in the United Kingdom 
by TJ International Ltd, Padstow, Cornwall 

The publisher's policy is to use permanent paper from mills that operate a 
sustainable forestry policy, and which has been manufactured from pulp 
processed using acid-free and elementary chlorine-free practices. Furthermore, 
the publisher ensures that the text paper and cover board used have met 
acceptable environmental accreditation standards. 

For further information on 

Blackwell Publishing, visit our website: 





Harvey Sacks, c. 1967 

Harvey Sacks, c. 1968 


Introduction by Emanuel A. Schegloff ix 

Note lxiii 

Acknowledgments lxv 

Part I Fall 1964-Spring 1965 

Lecture 1 Rules of conversational sequence 3 

Lecture 2 On suicide threats getting laughed off 12 

Lecture 3 The correction-invitation device 2 1 

Lecture 4 An impromptu survey of the literature 26 

Lecture 3 Suicide as a device for discovering if anybody cares 32 

Lecture 6 The MIR membership categorization device 40 

Lecture 7 On questions 49 

Lecture 8 On measuring 57 

Lecture 9 “I am nothing” 66 

Lecture 10 Accountable actions 72 

Lecture 1 1 On exchanging glances 8 1 

Lecture 12 Sequencing: Utterances, jokes, and questions 95 

Lecture 13 On proverbs 104 

Lecture 1 4 The inference-making machine 113 

Appendix A A Note on the Editing 126 

Part II Fall 1965 

[Lecture 1 and 2] [‘‘The baby cried. The mommy picked it up.”] 135 

Handout Group therapy session segment 136 

Lecture 3 A collaboratively built sentence; The use of ‘We’ 144 

Lecture 4 Tying rules 150 

Lecture 3 Tying rules; Insult sequences 157 

Lecture 6 ‘You’ 163 

Lecture 1 ‘Hotrodders’ as a revolutionary category 169 

Lecture 8 Invitations; Inexhaustable topics; Category-bound 

activities 175 

Lecture 9 Character appears on cue; Good grounds for 

an action 182 

Lecture 10 Clausal construction; Hotrodding as a test 188 

Lecture 1 1 Espousing a rule; Exemplary occurrences 193 



Lecture 12 ‘Tearing down;’ Non-translatable categories 199 

{Lecture 13] [‘Everyone has to lie’] 204 

Lecture 14 The navy pilot [from Sacks’ Research Notes] 205 

Appendix A “The baby cried” [Notes for lecture 1] 223 

Appendix B “The baby cried” [Notes on lecture 2] 230 

Part III Spring 1966 

Lecture 1 “The baby cried. The mommy picked it up” 236 

Lecture 1(R) “The baby cried. The mommy picked it up” 243 

Lecture 2 “The baby cried. The mommy picked it up” (ctd) 252 

Lecture 2(R) “The baby cried. The mommy picked it up” (ctd) 259 
[Lecture 3} [‘Everyone has to lie’] 267 

Handout Group therapy session segment 268 

Lecture 04.a An introduction sequence 281 

Lecture 04.b An introduction sequence (ctd) 292 

Lecture 4 Invitations; Identifications; Category-bound activities 300 

Lecture 5 Proffering identifications; The navy pilot; Slots; 

Paired objects, Adequate complete utterances 306 

Lecture 6 Omni-relevant devices; Cover identifications 312 

Lecture 7 Cover topics; Collaborative sentences; Tying 

rules; Relational-pair identifications 320 

Lecture 08 Orientation; Being ‘phoney;’ Hinting 328 

Lecture 8 ‘We;’ Category-bound activities 333 

{Lecture 9] [incorporated into lectures 8 and 10] 341 

Lecture 10 Pro-verbs; Performatives; Position markers; 

Warnings 342 

Lecture 11 ‘You’ 348 

Lecture 12 Warnings, challenges, and corrections; 

Explanations; Complaining-praising; Games 354 

Lecture 13 Button-button who’s got the button 363 

Lecture 14 Disorderability; Tying rules 370 

Lecture 1 5 Tying rules; Playing dumb; Correction-invitation 

device 376 

Lecture 16 Possessive pronouns; Possessables and possessitives 382 

Lecture 17 Pervasive, inexhaustable topics; Emblems 389 

Lecture 18 ‘Hotrodders’ as a revolutionary category 396 

Lecture 19 Appearance verbs 404 

Lecture 20 Character appears on cue; Good grounds for an action; 

An explanation is the explanation 410 

Lecture 21 Misidentification; Membership categories; Utterance 

pairs; Paradoxes 417 

/ Lecture 22] [combined with lecture 21] 427 

Lecture 23 Agreement; What can be done with language? 428 

Lecture 24 Measurement systems 435 

Lecture 23 ‘Company’ as an alternation category [incomplete] 44 1 


Lecture 26 Being ‘chicken’ versus ‘giving lip back’ 443 

Lecture 21 A mis-hearing (“a green?”); A taboo on hearing 450 

Lecture 28 Intelligibility; Causally efficacious categories 456 

Lecture 29 Place references; Weak and safe compliments 46 1 

Lecture 30 Various methodological issues 467 

Lecture 31 Games: legal and illegal actions 473 

Lecture 32 Seeing an ‘imitation’ 479 

Lecture 33 On sampling and subjectivity 483 

Appendix A ‘On some formal properties of children’s games’ 489 

Appendix B A Note on the Editing 507 

Part IV Winter 1967 

February 16 Omnirelevant devices; Settinged activities; ‘Indicator 

terms’ 515 

March 2 Turn-taking; Collaborative utterances via appendor 

questions; Instructions; Directed utterances 523 

March 9 Topic; Utterance placement; ‘Activity occupied’ 

phenomena; Formulations; Euphemisms 535 

Part V Spring 1967 

[Lecture 1—7} {‘One party speaks at a time’] 

Lecture 8 ‘‘Everyone has to lie” 549 

Lecture 9 ‘‘Everyone has to lie”(ctd) 557 

[Lecture 10] {incorporated into lectures 9 and 11] 567 

Lecture 1 1 ‘We;’ Category-bound activities; ‘Stereotypes’ 568 

Lecture 12 Category-bound activities; Programmatic relevance; 

Hinting; Being ‘phoney’ 578 

Lecture 13 Category-bound activities: ‘‘The baby cried;” Praising, 

warning, and challenging; Tautological proverbs 584 

Lecture 14 ‘Cover’ categories; Omni-relevance 590 

Lecture 13.1 ‘Safe’ compliments and complaints 597 

Lecture 13.2 Ultra-rich topics 601 

Lecture 16 Possessables and possessitives 605 

Lecture 1 7 Claiming possession; Emblems; Pro-terms and 

performatives; Utterance positioning 610 

[Lecture 18] {Paradoxes] 616 

Part VI Fall 1967 

General Introduction 619 

Lecture 1 The speaker sequencing problem 624 

Lecture 2 The ‘one-at-a-time’ rule; Violations; Complaints; 

Gossip 633 

Lecture 3 Bases for ‘interruption’ 64 1 



Lecture 4 Utterance completion; Co-producing an utterance; 

Appendor clauses 647 

Lecture 5 Utterance completion; Action sequencing; Appendor 

questions 656 

Lecture 6 Next-speaker selection techniques; Paired utterances 665 

Lecture 7 Intentional mis-address; Floor seekers 675 

Lecture 8 Pre-sequences 685 

Lecture 9 Paradoxes 693 

Lecture 10 ‘Everyone gets a chance to talk’ 701 

Lecture Oil Pronouns 711 

Lecture 11 Tying techniques 716 

Lecture 12 Repetition tying; “A green?” . . . “Who’s Wayne 

Morse?” 722 

Lecture 13 Tying-based mis-hearings; Locational tying; Pro-verbs 

and performatives 730 

Lecture 14 Paraphrasing; Alternative temporal references; 

Approximate and precise numbers; Laughter; ‘Uh huh’ 739 

Part VII Spring 1968 

April 17 Topic 752 

April 24 Second Stories 764 

May 8 Reason for a call; Tellability 773 

May 22 Pauses in spelling and numbering 784 

May 29 Verb selection; Interactionally generated invitations; 

Adequacy of local environments, etc. 787 

Appendix I ‘Introduction’ 1965 802 

Bibliography 806 

Index 813 


Introduction by Emanuel A. Schegloff ix 

Part I Fall 1968 

Lecture 1 

Second stories; “Mm hm;” Story prefaces; ‘Local news;’ 


Lecture 2 

Features of a recognizable ‘story;’ Story prefaces; 
Sequential locator terms; Lawful interruption 


Lecture 3 

Turn-taking; The notion ‘conversation;’ Noticeable 
absences; Greetings; Adjacency 


Lecture 4 

Turn-taking; Complaints about interruption; 


Lecture 3 

Collaboratives; Possible utterances; Utterance pairs; 
Greetings and introductions 


Lecture 6 

Greetings and introductions; Orientational utterances; 
Ultra rich, infinite topics; Being ‘phoney’ 


Part II 

Winter 1969 

Lecture 1 

Announcements; Touched-off utterances; Noticings; 
The makings of conversation; Local resources 


Lecture 2 

Safe Compliments 


Lecture 3 

‘Patients with observers’ as 'performers with audience’ 


Lecture 7 

Alternative sequences; Challenges; Claiming 


Lecture 8 

‘Identification reformulation;’ Pairing off at parties; 
‘Abstract’ versus ‘concrete’ formulations 


Lecture 9 

Sound shifts; Showing understanding; Dealing with 
‘utterance completion;’ Practical mysticism 



Verb uses; ‘A puzzle about pronouns’ 


Part III 

Winter 1970 

Lecture 1 

Foreshortened versus expanded greeting sequences; 
Voice recognition tests; Reason for a call; ‘My mind 
is with you;’ Tellability 


Lecture 2 

Conveying information; Story-connective techniques; 
Recognition-type descriptors; ‘First verbs;’ 
Understanding; Differential organization of perception 




Lecture 4 Greetings: Adjacency pairs; Sequential implicativeness; 

The integrative function of public tragedy 188 

Lecture 5 Foreshortened, normal, and expanded beginning 
sequences; Joking relationships; First topics; Close 
offerings 200 

Part IV Spring 1970 

Lecture 1 Doing ‘being ordinary’ 215 

Lecture 2 Stories take more than one utterance; Story prefaces 222 

Lecture 3 Story organization; Tellability; Coincidence, etc. 229 

Lecture 4 Storyteller as ‘witness;’ Entitlement to experience 242 

Lecture 5 ‘First’ and ‘second’ stories; Topical coherence; Storing 

and recalling experiences 249 

Lecture 6 Hypothetical second stories and explanations for first 
stories; Sound-related terms (Poetics); “What I didn’t 
do” 261 

Lecture 7 ‘What’s going on’ in a lay sense; Tracking 

co-participants; Context information; Pre-positioned 
laughter; Interpreting utterances not directed to one 269 

Lecture 8 Asking questions; Heckling 282 

Part V Winter 1971 

February 19 Poetics; Tracking co-participants; Touched-off topics, 

Stepwise topical movement 291 

March 4 Produced similarities in first and second stories; 

Poetics; ‘Fragile stories,’ etc. 303 

March 1 1 Poetics; Requests, offers, and threats; The ‘old man’ 

as an evolved natural object 318 

Part VI Spring 1971 

April 2 Introduction 335 

April 5 Poetics; Avoiding speaking first 340 

April 9 Technical competition 348 

April 12 Long sequences 354 

April 19 Caller-Called 360 

April 23 Characterizing an event 367 

April 26 An event as an institution 370 

April 30 Calling for help 376 

May 3 Problem solving; Recipient-designed solutions 384 

May 10 Agent-client interaction 391 

May 1 7 Poetics: Spatialized characterizations 396 

May 21 Closing; Communicating a feeling; Doctor as 

‘stranger’ 402 

May 24 “Uh huh;” Questioner-preferred answers 410 

Contents vii 

Part VII Fall 1971 

Lecture 1 On hypothetical data; Puns; Proverbial expressions 419 

Lecture 2 Doing ‘understanding;’ Puns 425 

Lecture 3 Allusive talk; Poetics 431 

Lecture 4 Spouse talk 437 

Lecture 3 Selecting identifications 444 

Lecture 6 A ‘defensively designed’ story 453 

Lecture 7 The ‘motive power’ of a story; ‘Ex-relationals’ 458 

Lecture 8 Preserving and transmitting knowledge via stories 466 

Lecture 9 The dirty joke as a technical object; Temporal and 

sequential organization; ‘Guiding’ recipient 470 

Lecture 10 The dirty joke as a technical object (ctd); Suspending 

disbelief; ‘Guiding’ recipient; Punchlines 478 

Lecture 1 1 The dirty joke as a technical object (ctd); Packaging 

and transmitting experiences 483 

Lecture 12 The dirty joke as a technical object (ctd); “What is sex 

like;” Possible versus actual applicability of a rule 489 

Lecture 13 Two ‘floor-seizure’ techniques: Appositional expletives 

and “Uh” 495 

Lecture 14 The working of a list; Doing ‘hostility’ 499 

Lecture 13 ‘Fragile’ stories; On being ‘rational’ 504 

Lecture 16 On dreams 512 

Part VIII Spring 1972 

Lecture 1 Adjacency pairs: Scope of operation 521 

Lecture 2 Adjacency pairs: Distribution in conversation; A single 

instance of a Q-A pair 533 

Lecture 3 A single instance of a phone-call opening; 

Caller-Called, etc. 542 

Lecture 4 The relating power of adjacency; Next position 554 

Lecture 3 A single instance of a Q-A pair; Topical versus pair 

organization; Disaster talk 561 

Lecture (6) Laughing together; Expressions of sorrow and joy 570 

References 576 




The publication of these lectures makes publicly available virtually all of the 
lectures by Harvey Sacks on conversation and related topics in social science. 
Most of the lectures in this larger corpus were originally delivered to classes at 
the University of California - first to sociology classes at the UCLA campus, 
and then (beginning in Fall 1968) to classes in the School of Social Science at 
the Irvine campus of the University. 

Although Sacks produced copious analytic notes, many of which served as 
materials for his lectures, what is presented here are the lectures themselves, 
transcribed from tape recordings. Almost all of Sacks’ lectures were initially 
transcribed by Gail Jefferson, although most of the material for Fall 
1964-Spring 1965, in that it antedates either her contact with Sacks and this 
work, or her undertaking to transcribe the lectures, was intially transcribed by 
others. With one exception (Sacks, 1987 {1973}), it is also Jefferson who has 
edited those lectures which have previously been published, as well as the 
lectures published here. 1 As noted in her introductory notes to the several 

My thanks to Paul Drew, John Heritage, Gail Jefferson, Michael Moerman and 
Melvin Pollner for sensitive responses to a draft of an earlier version of part of this 
introduction (prepared for the 1989 publication of the 1964-5 lectures), and for 
suggestions which I have in some cases adopted without further acknowledgement. 
I am further indebted to John Heritage and Michael Moerman for their generous and 
helpful comments on a draft of the present introduction/memoir, and to Gail 
Jefferson for calling to my attention what she took to be lapses in accuracy or 

1 Of the lectures published here, the set for 1964-5 were published in a special 
issue of the journal Human Studies , 12, 3-4 (1989), and of those, the following had 
been previously published elsewhere, edited by Gail Jefferson: 

Fall 1964-5, lecture 5 has been published under the title ‘You want to find out 
if anybody really does care’ in Button and Lee (1987: 217-25). 

Winter 1964-5, lecture 14 has been published under the title ‘The inference 
making machine: notes on observability’ in van Dijk (1985: 13-22). 

Other than the 1964-5 lectures, the following lectures have been previously 
published, also edited by Jefferson: 

Spring 1966, lecture 18 (and related material in Fall 1965, lecture 7), under the 
title ‘Hotrodder: a revolutionary category,’ in Psathas (1979). 

Spring 1966, lecture 13, under the title ‘Button-button who’s got the button,’ in 
Zimmerman and West (1980: 318-27). 

Spring 1966, lecture 24 (with excerpts from Fall 1967, lecture 14; Winter 1970, 
lecture 2; and Spring 1970, lecture 3), under the title ‘On members’ measurement 
systems,’ (Sacks, 1988/89). 




‘lectures’ and in Appendix II in her editor’s notes to the previous publication 
of the 1964-5 lectures in the journal Human Studies , those lecture-texts have 
been pieced together from several sets of lectures which Sacks gave during the 
1964-5 academic year, to make a more coherent and readable document. 
These early ‘sets’ of lectures are full of gaps, and it is not always clear just 
when some lecture was given. Accordingly, the reader should bear in mind 
that this presentation of Sacks’ early lectures cannot be used to track the 
development of themes over time, to trace what topics or themes appear to 
have been related in Sacks’ thinking, etc. 

Otherwise, it needs to be said at the outset with respect to the present 
edition that the editorial undertaking has been monumental and its execution 
heroic. This the reader can only partially see, for what has not been included 
is, for that reason, not apparent, nor is the work of sorting and collating what 
is made available in these volumes. This work has, as a matter of course, 
involved divergences of several kinds from the texts of these lectures which 
have circulated in various forms of reproduction over the years. These are 
largely stylistic in nature, and are clearly designed to render the text more 
accessible, more readable, and more consistent in stance, point of view, diction, 
etc. 2 On occasion, however, these textual adjustments could be misread as 
taking a stand on an analytic matter which Sacks otherwise addresses, could be 
given a ‘political’ reading, or could appear to have a ‘political’ upshot, and it 
would be well for the reader to be alerted to such possibilities. 

By ‘taking a stand on an analytic matter which Sacks otherwise addresses’ 
I mean to call attention to such adjustments in diction as one in which Sacks 
follows an excerpt from a group therapy session by referring to one of the 
speakers as “this fellow Dan” (in the originally circulated transcript of the 
lecture), a reference which is in the present edition rendered as ‘the therapist.’ 
Sacks takes up the issue of the description of persons, and category-ascriptions 
such as ‘therapist/patient,’ on several different occasions in these lectures and 
in several papers. Because of the options available for formulating persons, 
particular choices of descriptors or identification terms served, in Sacks’ view, 
to pose problems for analysis, and could not properly be invoked or employed 
in an unexamined way. Accordingly, no particular claims should be under- 

Winter 1970, lectures 1 and 2, under the title ‘Some considerations of a story told 
in ordinary conversations,’ (Sacks, 1986). 

Spring 1970, lecture 1 (with excerpts from Winter 1970, lecture 2; Spring 1970, 
lecture 4; and Spring 1971, lecture 1), under the title ‘On doing being ordinary,’ in 
Atkinson and Heritage (1984). 

Fall 1971, lectures 9-12, under the title ‘Some technical considerations of a dirty 
joke,’ in Schenkein (1978). 

In addition, extracts from a number of lectures have been assembled by Jefferson 
as ‘Notes on methodology,’ in Atkinson and Heritage (1984: 21-7). 

2 Cf. the editor’s notes by Gail Jefferson (Jefferson, 1989), and, in the present 
edition, footnotes at Fall 1964-Spring 1965, lecture 2, p. 18; Spring 1966, lecture 
04. a, p. 281; Winter 1969, lecture 7, p. 120; as well as Appendix A to lectures for 
Fall 1964-Spring 1965, and Appendix B to lectures for Spring 1966. 



stood as implied by occasional references to participants by such category 
terms in the current text (cf. the editor’s Appendix A, Spring 1966). 

By ‘political’ I mean, in this context, a relative positioning by Sacks of 
himself, his undertaking, his colleagues, his students, other contemporary 
intellectual undertakings, the established contours of the disciplines (sociol- 
ogy, linguistics, anthropology, etc.) and their groupings (e.g., the social 
sciences), the physically present class to which he was ostensibly addressing 
himself 3 and the like. Deployment of the pronouns ‘we,’ ‘you,’ ‘they’ and the 
like can serve to express varying sorts of solidarity and differentiation, and 
different ways of ‘partitioning the population’ (as he used to put it). 4 This was 
a matter to which Sacks was sensitive, having written a paper in graduate 
school only a few years earlier on Durkheim’s use of ‘we’ in The Elementary 
Forms of the Religious Life, an echo of which appears in lecture 33 of the 
Spring 1966 lectures. Where the text suggests such alignments, readers 
should exercise caution. 

It must also be recalled that the omission of some lecture sets in the present 
edition, and the transposition of some lectures from one set to another, 
requires that caution be used in basing an analysis of the appearance and 
development of themes, ideas and discussions of data fragments on this 
edition alone. The full texts of prior versions of all the lectures will be 
available through the Sacks Archive at the Department of Special Collections 
of the UCLA Library. 

These cautions aside, it should be said that one cannot really retrieve Sacks’ 
‘voice’ from the text as presented here. In the interest of readability and of the 
accessibility of the content, what was sometimes a real challenge to discursive 
parsing - even to his closest friends and colleagues - has been smoothed out. 
Gone are the often convoluted phrasing, the syntax that might or might not 
come together at the end, the often apparently pointillistic movement from 
observation to observation - sometimes dovetailing at the end into a coherent 
argument or picture, sometimes not. The very long silences, of course, were 
lost in the transcribing process. 

But Sacks himself treated his habits and manners, his attitudes and 
convictions, as ‘private’ (as he puts it in response to a question as to whether 
he is ‘convinced’ that single events are studyable after the general introductory 
lecture, Fall 1967, “That’s such a private question”), and not really relevant 

3 Cf. the lecture of April 2 in the Spring 1971 lectures, on Sacks’ notion that he 
was really talking to colleagues, friends and ‘students’ wherever they might be who 
were interested in his current work and not necessarily to the class actually in the 

1 For example, in lecture 6 of the 1964-5 lectures an alignment may seem to be 
implied in which Sacks identifies himself with the physically present students in 
criticism of “all the sociology u>e read,” whereas the text of the lecture as previously 
circulated had read “all the sociology you read ...” (emphases supplied). 

See Appendix A to the Fall 1964-Spring 1965 lectures for the editor’s account of 
Sacks’ use of personal pronouns such as ‘you,’ ‘I’ and ‘we’ in the lectures, and of her 
editorial practices for changing some of these references in preparing this edition. 



to what he felt merited the attention of others in what he had to say. It is that 
which these volumes present. As quickly becomes apparent from the texts of 
the lectures, we have yet to take the full measure of the man. 

These series of lectures present a most remarkable, inventive and produc- 
tive account of a strikingly new vision of how to study human sociality. With 
but a few exceptions, the students who sat in the rooms in which the lectures 
were delivered can hardly have known what they were hearing. The lectures 
were addressed to non-present students, to those who might come to know 
what to make of them. That audience continues to grow. 

Under what circumstances were these lectures delivered and recorded? 
What is their intellectual and scientific context? What is most notable in 
them? These matters cannot be dealt with comprehensively here, but a brief 
treatment, in a mixed genre which might be termed an ‘introduction/ 
memoir,’ can help provide an overview and some setting for what is 
increasingly recognized as a startlingly original and important address to the 
social organization of mind, culture and interaction. 


Sacks received his AB degree in 1955 after three years at Columbia College. 
In later years, Sacks would reminisce with partly feigned and partly genuine 
awe about the faculty at Columbia - Jacques Barzun, Meyer Schapiro, Lionel 
Trilling, various students and former students of Franz Neumann such as 
Julian Franklin and Peter Gay (and Neumann himself, who, however, may 
well have not been teaching undergraduates when Sacks was there), although 
it was never entirely clear with which of these ‘eminences’ Sacks had himself 
studied. 5 

Although he did not officially ‘major’ in sociology, Sacks’ education was 
influenced in an important way by C. Wright Mills. The influence was not 
channeled primarily through course work; most important to Sacks was that 
Mills secured for him a faculty-authorized access to the stacks of Butler 
Library and turned him loose on his own. But Sacks would later say that from 
Mills he had learned ‘audacity.’ 

In spite of the predominantly socio-cultural cast of the faculty who figured 
most centrally in Sacks’ later reminiscences, the two closest college friends 
with whom Sacks kept in touch later on were both biologists. 

Upon graduation from Columbia, Sacks was awarded a scholarship at Yale 
Law School where he earned his LLB in 1959- While at Yale, he participated 
in the group around Harold Lasswell, and became more interested in 
understanding how the law as an institution worked, how it could work, than 

5 I recall an account of how students would celebrate if they achieved grades of ‘A’ 
from Trilling or Schapiro, but it was unclear, at least to me, whether Sacks himself 
had been one of those students. 



in making it work as an attorney himself. 6 He went looking for intellectual 
resources with which to pursue this interest and turned first to Cambridge, 
and to the work of Talcott Parsons in particular (although formally he was 
enrolled as a graduate student in Political Science at MIT, and was employed 
as a research assistant in the Department of Economics and Social Sciences 
there). But what he found in Cambridge that was most consequential for the 
subsequent development of his thinking was not Parsons (and not Chomsky, 
some of whose lectures at MIT he attended), but rather Harold Garfinkel. 

Garfinkel was spending a sabbatical leave from UCLA at Harvard, where 
he had himself earned his Ph D a number of years earlier. Sacks and Garfinkel 
met at Parsons’ seminar in Cambridge, and were immediately attracted by 
each other’s seriousness. Their intellectual relationship was sustained until the 
early 1970s. However, in 1959-60, when it became clear to Sacks that the 
solutions to the problems he had set himself were not to be found in 
Cambridge, he followed his law school teacher Lasswell’s advice, and decided 
to pursue graduate work in sociology at Berkeley. 

Berkeley appealed on several grounds. Laswell had suggested that Sacks 
pursue his interests through the continuing study of labor law and industrial 
relations. An attractive locale was furnished by the Institute of Industrial 

6 Sacks once recounted a story which provides some insight into the appeal which 
Garfinkel’s work must have had for him when he later encountered it. 

He was engaged in a discussion with several other law school students arguing 
through some problem in case law which they had been set - a problem in torts, if 
I remember correctly. The issue was whether or not a person on the ground was 
entitled to recover damages incurred from the overflight of his property by an 
airplane. At one point, in a kind of mimicry of the ‘how many hairs make a bald 
man’ paradox, the students coped with the argument that no damages could be 
collected if the plane was being piloted in a proper and accepted manner by seeing 
how far they could press the definition of what was ‘proper.’ What if it were flying 
at 2,000 feet? At 1,000 feet? At 250 feet? At 5 feet? Sacks reported that when the 
last of these proposals was offered, it was dismissed as ‘unreasonable,’ as frivolous, as 
violating the canons of ‘common sense.’ But, he pointed out, that could as well have 
been said about the penultimate one, but wasn’t. What struck him, then, and 
puzzled him, was that the 'legal reasoning’ which was the much heralded instrument 
in whose use they (students of the law) were being trained rested on, and was 
constrained by, an infrastructure of so-called ‘common sense’ which was entirely tacit 
and beyond the reach of argument, while controlling it. And, in that legal reasoning 
was something on which the entire legal structure rested (and not just particular areas, 
such as torts, contracts, crimes, etc.), how the law as an institution actually worked, 
what made it work the way it did, what restrained its reasoning from pressing the law 
in other directions, was shrouded in mystery. Undoubtedly, this was only one of the 
puzzles about how the law could work which engaged Sacks’ interest, but it is one 
for the solution of which Garfinkel’s work on methods of commonsense reasoning 
and practical theorizing, then in progress, would have been an attractive resource. 

The issue prompted by this law school incident gets articulated explicitly for its 
bearing on working with recorded conversational materials at the beginning of lecture 
1 for Fall 1971; cf. volume 2 of the present edition. 



Relations at Berkeley, and in particular by Philip Selznick whose interest in 
organizations and bureaucracy was complemented by a developing interest in 
legal institutions. (Indeed, several years later Selznick was to establish the 
Center for the Study of Law and Society at Berkeley, and Sacks was to be 
among its first graduate fellows.) But Berkeley was attractive on other counts 
as well. Aside from its having developed one of the strongest sociology 
departments in the country, Sacks was attracted by the presence of Herbert 
Blumer, whose SSRC monographic critique (1939) of Thomas and 
Znaniecki’s Polish Peasant in Europe and America Sacks had found penetrat- 
ing. (Sacks lost interest in Blumer soon after arriving in Berkeley, and did not 
study with him at all.) 

It is worth pausing a moment to recall where some of the relevant 
American social sciences stood during these formative years of the late 1950s 
and early 1960s, at least as they appeared to graduate students, to some 
graduate students, to the graduate students who figure in this account. 

There had not yet been the rise to professional visibility of a radical 
sociology. C. Wright Mills’ Sociological Imagination was still a daring 
manifesto, his Power Elite still a model inquiry. Theory was predominantly (as 
it was then called) ‘structural-functionalist’ and especially Parsonian. ‘Empir- 
ical’ sociology was still predominantly ‘Columbia-oriented’ rather than 
‘Chicago-oriented;’ data analysis was multivariate, not regression-based. Blau 
and Duncan’s The American Occupational Structure was still half a decade to 
a decade away. And social psychology was in large measure a choice between 
‘small groups’ of the Bales variety or of the Michigan group dynamics variety, 
a substantial dollop of ‘public opinion’ or ‘attitudes’ research, with a minority 
voice somehow identified - often wrongly - with symbolic interactionism: 
Blumer at Berkeley being the most visible - or vocal - representative, Goffman 
(‘ The Presentation of Self in Everyday Life, published in the United States in 
1959) just beginning to be recognized, Becker still largely unknown. 

In anthropology, the Gumperz/Hymes special issue of the American 
Anthropologist was not to appear until 1964, ethnoscience and componential 
analysis were just coming into their own, the ethnography of communication 
was just beginning to recruit its hoped-for army to canvas the world. 

In linguistics, Chomsky’s Aspects of the Theory of Syntax was not published 
until 1964, outsiders were just registering the import and impact of his review 
of Skinner’s Verbal Behavior and his Syntactic Structures (1957). Linguistics 
was just beginning to establish a track record for its significance to other 

Throughout his stay in Berkeley, Sacks remained in touch with Harold 
Garfinkel (now returned to his home base at UCLA) whose program of 
ethnomethodological studies was being developed in a series of writings which 
were privately circulated for the most part in mimeographed form. (It should 
be recalled that it was not until 1959 that Garfinkel’ s ‘Aspects of the problem 
of commonsense knowledge of social structure’ was published - and not in a 
broadly accessible outlet at that; not until I960 for ‘The rational properties of 
scientific and commonsense activities,’ also not in a source generally read by 



sociologists; not until 1964 that ‘Studies of the routine grounds of everyday 
activities’ appeared in Social Problems-, and not until 1967 that Studies in 
Etbnomethodology was published.) It was largely through Sacks that these 
manuscripts came to be circulated in Berkeley, largely among graduate 
students in sociology. Of course, Sacks did not only circulate Garfinkel’s 
manuscripts; in discussions among the students he added the special directions 
of his own thinking, in some respects converging with Garfinkel’s, in other 
respects quite distinctive. 

At the time, Garfinkel was co-principal investigator with Edward Rose of 
the University of Colorado on a research grant which supported a series of 
conferences in Los Angeles in which Sacks took part. So Sacks’ engagement 
with Garfinkel’s manuscripts in northern California was complemented by 
more direct, personal engagement in the south. At the same time, other 
developments were in progress in both north and south; in the north, for 
example, Selznick had brought into his new Center for the Study of Law and 
Society a number of graduate students in the social sciences, and especially 
sociology. During the 1962-3 academic year, this group included Sacks, 
David Sudnow and the present writer, whose activities separately and 
together were to contribute to future developments, but are not directly in 
point here. 

In 1963, Garfinkel arranged for Sacks to move to Los Angeles. He was to 
have an appointment as Acting Assistant Professor of Sociology at UCLA, 
with the first year off. During that year, 1963-4, Garfinkel and Sacks 7 were 
to serve as Fellows at the Center for the Scientific Study of Suicide in Los 
Angeles, under the sponsorship of its director, Edwin Schneidemann. As it 
happens, my own work prompted a move from Berkeley to Los Angeles 
during the summer of 1963, and Sacks and I continued both a work and a 
personal relationship during that year. I can therefore describe, at least in brief 
compass, his primary intellectual preoccupations during the year. A great 
many of them had his involvement with the Suicide Prevention Center as a 
point of departure, thereafter taking the often surprising directions which his 
distinctive mind imparted to them. In diverse ways, these interests show up 
in his first ventures in teaching, the 1964-5 lectures which provided the point 
of departure for the further development of the work, presented in the 
subsequent lecture series published here. 

One line of these concerns focussed on an examination of psychiatric, and 
especially psychodynamic, theorizing, which furnished one primary theoreti- 
cal handle on the phenomenon of suicide at the SPC, and which, more 
particularly, was key to the so-called ‘psychological autopsies’ which were 
conducted following suicides and which were of very great interest to both 
Sacks and Garfinkel. Thinking about psychodynamic theorizing led Sacks (as 
it had led me; cf. Berkeley Journal of Sociology, 1963) to a concern with 
dialogue, and in particular with Platonic dialogue as a form of discourse 
designed to control conduct. That, in turn, led him to a more general interest 

7 And Erving Goffman, visiting on an occasional basis from Berkeley. 



in Greek philosophy, and particularly in Greek logic (on which he was 
reading, among other sources, Kneale and Kneale’s The Development of Logic, 

From the Freudian theorizing, from a prior interest in ‘children’s cultures’ 
set off by the work of the Opies (1959), and from a persistent attention to the 
problems posed by the apparent facts and achievements of socialization, there 
developed an interest in the behavior of children. This interest Sacks pursued 
largely through examination of source books on children’s games (an interest 
prompted as well by the work of O. K. Moore on games as ‘autotelic folk 
models;’ cf. Anderson and Moore, I960), of the studies and protocols of 
Barker and Wright (1951, 1954), in observational studies Sacks acquired 
from Roger Brown, and other sources. 

And Sacks pursued a number of other scholarly interests, in biblical studies 
and interpretation, in translation, in archaeology, etc. In a very different vein, 
Sacks came across stenographic transcripts, and then the tapes, of the 
telephone calls to the Suicide Prevention Center of, or about, suicidal persons. 
All of these themes may be found in the 1964-5 lectures, but it was the last 
of them which provided the proximate source for the focussed attention on 
talk itself - perhaps the most critical step toward the development of 
conversation analysis. 

Throughout the 1963-4 academic year, Sacks and I continued the 
discussions and explorations entered into in Berkeley during the preceding 
year and a half. This is not the place for a substantial account of those 
activities (on-site explorations of the possibilities of field observation at the Los 
Angeles International Airport, in the reference room of the UCLA library, at 
neighborhood ‘Okie’ bars in Venice, and elsewhere; long discussions on the 
UCLA campus where I was a visiting scholar, at the beach in Venice where 
he lived, or at the apartment at the fringe of Beverly Hills where my wife and 
I lived). But it may be of interest to describe what seemed to me at the time 
something quite new, and seems to me now in retrospect the first appearance 
of what would eventually become, after a number of major transformations, 
what is now called ‘conversation analysis.’ 

It was during a long talking walk in the late winter of 1964 that Sacks 
mentioned to me a ‘wild’ possibility that had occurred to him. He had 
previously told me about a recurrent and much discussed practical problem 
faced by those who answered phone calls to the Suicide Prevention Center by 
suicidal persons or about them - the problem of getting the callers to give 
their names. Now he told me about one call he had seen/heard which began 
something like this: 

A: This is Mr Smith, may I help you. 

B: I can’t hear you. 

A: This is Mr Smith . 

B: Smith. 

After which Mr Smith goes on, without getting the caller’s name. And later, 

Introduction xvii 

when Mr Smith asks for the caller’s name, the caller resists giving it. On the 
one hand, Sacks noted, it appears that if the name is not forthcoming at the 
start it may prove problematic to get. On the other hand, overt requests for 
it may be resisted. Then he remarked: Is it possible that the caller’s declared 
problem in hearing is a methodical way of avoiding giving one’s name in 
response to the other’s having done so? Could talk be organized at that level 
of detail? And in so designed a manner? 

A month or two later, I arrived home at our apartment in the late 
afternoon, to find Sacks waiting for me there. A transient difficulty with 
Garfinkel had led him to realize that, if not on the present occasion then at 
some future time, he might have to fend for himself in the academic 
marketplace and had better have some written work to show. So he had 
drafted the sketches of two papers. I left him talking to my wife in the living 
room and retreated to my study and read the sketches. One of them was 
about a methodical way of avoiding giving one’s name. As the reader who 
turns to the 1964-5 lectures will soon discover, this is where Sacks’ lectures 
began (not only in the composite version assembled for this publication, but 
in the original as well). 

Why might this episode, and these observations, be treated as the 
beginning of what would come to be called ‘conversation analysis’? 8 Because 

8 In the ‘General Introduction’ lecture for Fall 1967, (p. 621), Sacks introduces 
the work to be presented by describing “When I started to do research in 
sociology ...” It is unclear what Sacks means to refer to: when he went to 
Cambridge? to Berkeley? sometime during graduate school? in Los Angeles? are these 
the right terms to locate the reference? 

In a way, the 1963 paper ‘Sociological description’ is not incompatible with the 
account offered in this Fall 1967 lecture, except for the description (p. 622) of 
starting “to play around with tape recorded conversations,” which surely did not 
happen until the year at the Suicide Prevention Center. Until then, friends of Sacks 
will remember occasions of sitting ‘with him’ in some public place and suddenly 
realizing that Sacks was no longer in the same interaction, but was overhearing a 
nearby conversation, and often taking out the omnipresent little multi-ring notebook 
and jotting down a fragment of the talk and some observations about it. The virtues 
of “replayling] them . . . type[ing] them out somewhat, and studying] them 
extendedly” (Fall 1967, ibid.) were realized against a long experience of such 
overhearing and notetaking. (One shared experience which may have alerted Sacks to 
the payoffs of taking materials like the SPC tapes seriously was my experience during 
1962-3 in Berkeley at the Law and Society Center of tape recording psychiatric 
competency and criminal insanity examinations for subsequent analysis.) 

But it is worth noting that Sacks did not set out to study conversation or language 
in particular. His concern was with how ordinary activities get done methodically and 
reproducibly, and the organization of commonsense theorizing and conduct which 
was relevant to those enterprises. Clearly, he found talk, or what was being done 
through talk, of interest before coming upon taped materials - else he would not 
have been jotting overheard bits in notebooks. But the taped material had clear 
attractions when it became available as a resource, and the talk invited being dealt 
with as an activity in its own right. But that was something that turned out from 
experience, not something that had been aimed at, or ‘theoretically projected.’ 

xviii Introduction 

there is the distinctive and utterly critical recognition here that the talk can be 
examined as an object in its own right, and not merely as a screen on which 
are projected other processes, whether Balesian system problems or Schutzian 
interpretive strategies, or Garfinkelian commonsense methods. The talk itself 
was the action, and previously unsuspected details were critical resources in 
what was getting done in and by the talk; and all this in naturally occurring 
events, in no way manipulated to allow the study of them. And it seemed 
possible to give quite well-defined, quite precise accounts of how what was 
getting done was getting done - methodical accounts of action. 

This was just the start of a long train of quite new things that Sacks was 
to provide. It was only a little over a year later that the eventually published 
version of ‘An initial investigation . . .’ (1972a) was completed. It is hard 
now to appreciate how startlingly new and unprecedented that paper was at 
the time. If one recalls the publication history of Garfinkel’s work (and that 
GofFman’s Behavior in Public Places was published in 1963, and Relations in 
Public was not to be published until 1971), a sense of its uniqueness when it 
was published in 1972 might be somewhat more accessible. Its utter 
originality in 1964-5 when it was being written, and the originality of the 
materials in the first of these lectures which were delivered around the same 
time, may be better grasped by reference to this other work. With the current 
wisdom of hindsight, of course, our sense of this early work of Sacks’ is readily 
assimilated to the direction we now know such studies took. But the 
originality was not only startling in 1964 and 1965; it had the additional 
headiness — and vertigo - of indeterminateness: How might one proceed? 
What sort of discipline was this or might it be? Once a previous sense of 
plausibility about the depth and detail of organization in conduct and 
apperception of the world were set aside, what constraints on inquiry were 
defensible? To what level of detail was it sensible to press? 

During the summer of 1964, I left Los Angeles for the mid-west, 
wondering what ever Sacks would do about lecturing to UCLA undergrad- 
uates, and wondering as well how our contact could be sustained. The latter 
problem was solved in part by a variety of resources that allowed me recurrent 
trips to California during the 1964-5 year (though less so in ensuing years), 
and in part by a practice which also satisfied my curiosity in the first respect. 
Sacks would tape record his lectures and send them to me, and (if I remember 
correctly) to David Sudnow who was spending the year in St. Louis, doing the 
field work for his dissertation, later to appear as the book Passing On (1967). 
At irregular intervals I would receive in the mail a little orange box with a 
yellow label, containing a three-inch reel of tape, enough for the 50-minute 
lectures (more or less) which Sacks was delivering. The lectures were for me, 
then, a rather special form of monologic telephone call interspersed with our 
dialogic ones (which were not recorded), and then, after Gail Jefferson started 
transcribing the lectures, they were a sort of long letter series. 9 It turns out that 

9 At the time they were being delivered, I encountered the lectures term by term, 
like long analytical letters from Sacks. I had little overall view of them and of their 

Introduction xix 

they became Sacks’ most successful and prolific form of scientific communi- 

When he wrote papers, Sacks imposed standards of formality and precision 
that were extremely hard for him to meet to his own satisfaction. Most of the 
papers he published under his own name alone were work-ups of lectures . 10 
Most of the papers he drafted on his own as papers he was never sufficiently 
satisfied with to publish. The exceptions, ‘An initial investigation . . . ’ or ‘On 
some puns, with some intimations’ give some idea of what Sacks thought a 
finished piece of work might look like. 

Aside, then, from his collaborative publications, the lectures are the vehicle 
by which most of his work was made available. Perhaps it was the explicitly 
and necessarily informal and limited character of the occasion that could allow 
him to get ‘the stuff’ out the best he could, with no pretense to finally 
getting it ‘just right.’ Those who have seen some of his successive versions 
of the ‘same pieces’ will know how great a change could overtake some 
piece of work under the guise of getting it just right . 1 1 But the quality of 
what was delivered in those lectures, and in those which followed, and the 

overall development, of long term changes in the work reflected in them, etc. This 
was largely because such changes would have come up in, or (without necessarily 
being explicitly discussed) informed, our conversations with each other in the interim 
between shipments, or could not be recognized for the changes they represented until 
later developments. Largely, then, my reading was marked by my being struck, 
charmed, and often amazed at what Sacks’ sleight of hand could materialize out of 
a bit of data, the twist he could impart - no, discover - in it, the tacit understandings 
he could, by a flash of insight, show we (‘casual’ readers or onlookers) had furnished 
it. Sometimes the ‘twist’ assumed the proportions of a whole analytic topical area - 
e.g., storytelling structure. I came to the reading of each new ‘package’ with a kind 
of avid curiousity about what sorts of new things - whether unexpected observations 
about a moment or whole new analytic issues - were tucked into those pages, and the 
reading proceeded from flash to flash. It was like watching one’s athletic friend show 
what he could do. 

Preparation of this publication and this introduction has afforded me the occasion 
for a larger overview, or series of overviews - of each set of lectures and of the set of 
sets. In them I am brought to recall or to discover in retrospect larger scale 
movements and changes, emerging and waning themes. Of course, this is refracted 
through my own experience and intellectual colleagueship with Sacks. I have tried to 
strike a balance between that kind of perspectival account and a less personalized 
overview and setting-into-context. 

10 Cf. for example, the paper on story-telling (1974). The paper on puns (1973) 
is an exception here, having never been fully worked up as a lecture before being 
prepared for the Georgetown Round Table, in whose proceedings it was published. 
‘Everyone has to lie’ (1975) was adapted from a lecture, but the materials for the 
lecture were initially drafted as a paper, under the title ‘The diagnosis of depression,’ 
which was never published in its original ‘paper’ format. 

11 See the initial two lectures of the Spring 1966 term presented in this edition 
with Sacks’ first effort at revision, at pp. 236-46 below, for a sample. In this case, 
a virtually identical version of the same material was eventually published as ‘On the 
analyzability of stories by children’ (Sacks, 1972b). 

xx Introduction 

special vision that underlies it, did not require getting it ‘just right’ to be 
apparent. 12 

Although he continued to tape a variety of research and teaching activities, 
Sacks stopped recording his lectures in 1972 for a number of reasons. Some 
of his lectures at the Linguistic Institute of 1973 at the University of Michigan 
were recorded, as were some of the sessions of the joint seminar we taught, but 
these were not recorded by Sacks, and were not reviewed for transcription by 

Harvey Sacks was killed in an automobile accident in November 1975 
while on his way to the campus of the University of California, Irvine where 
we were to meet to formulate a program which we were discussing 
establishing at the Santa Barbara campus of the University. One can hardly 
imagine what the next years of Sacks’ intellectual life would have produced, 
especially in an academic environment fully supportive of the enterprise which 
had already developed. 


The ‘first installment’ of these lectures - the ones delivered during the 
1964-5 academic year - can be furnished with two sorts of intellectual 
reference points - ones in Sacks’ own intellectual development and ones in the 
intellectual context around him. 

In his own thinking, these lectures come after his paper ‘Sociological 
description’ (1963), written in 1962-3 in Berkeley 13 and during the same 
period as ‘An initial investigation . . (1972a) which was finished in June 

1965. 14 

Several features of these early papers which serve as landmarks in Sacks’ 
intellectual terrain, and of the early lectures, display some of the most potent 
influences on his thinking at that time. There is first of all a wide-ranging 
responsiveness to Garfinkel’s thematics, broadly acknowledged in a footnote 
to ‘Sociological description’ (1963: 1), and in recurrent notes in the early 
writings and lectures. A thorough treatment of the influences here, I daresay 
the reciprocal influences at work here, remains to be written. At a different 
level, there is the transparent allusion to the later Wittgenstein embodied in 

12 Still, readers should bear in mind the in-progress status which this work had for 
Sacks. While still alive, he expressed a willingness to have the lectures published, if 
the publication could be done without much editing, not only because he did not 
want to spend the time, but also to avoid masking the work-in-progress nature and 
status of the effort. It should be a way of getting ‘a lot of stuff’ noticed, without 
suggesting what should in the end be fashioned from it. The lectures were not meant 
to look finished. 

13 See the discussion below of the Spring 1966 lectures, and of ‘possible 
description’ in particular. 

14 Cf. the initial footnote to the version published in Sudnow (1972b). 



the invention (ibid.) of the ‘commentator machine’ as a grand metaphor for 
(variously) the relationship of social science discourse to the social world which 
is its object, of commonsense or lay talk about the world to ordinary 
enactments of it, etc. 

Perhaps less expectable in the contemporary academic setting, in which 
studies of discourse and conversation are often set in contrast to transforma- 
tional grammar, is the echo of generativist studies in the form of some of this 
early work, and especially in the form of its problem development. Take as 
a case in point ‘On the analyzability of stories by children’ (1972b, revised 
from the first two lectures for Spring 1966, but first worked up as lectures 1 
and 2 for Fall 1965). 

The data for that set of lectures and publication, it will be recalled, were 
taken not from ordinary conversation, but from the response of a young child 
to a request by an adult for a story. Most relevantly for the present discussion, 
this had the consequence that there was no ensuing talk by a co-participant 
which could be examined to reveal an understanding of the ‘story’ which was 
‘indigenous’ to the interaction, along the lines exploited in later conversation- 
analytic work. In its place, Sacks relied on his understanding of the text being 
examined (“The baby cried. The mommy picked it up”), and the under- 
standing which he attributed to his audience - understandings not overtly 
provided for by the text itself (for example, that ‘the mommy’ is the mommy 
of that baby, although the story as told by the child was expressed as ‘ the 
mommy picked it up’). 

The problem, as Sacks developed it, was to build ‘an apparatus’ that would 
provide for such hearings or understandings, and would serve both as a 
constraint on them and as a research product to which they could lead. This 
form of problematics, of course, echoes the commitment to build a syntactic 
apparatus which would provide for the alternative parsings of a claimedly 
ambiguous sentence such as ‘Flying planes can be dangerous’ (Chomsky, 
1957). The reader is first asked to recognize that alternative ‘structural 
interpretations’ can be assigned to this sentence, and then to be concerned with 
the construction of a syntax that produces such an ambiguity and provides for 
its disambiguation. To be sure, this form of problem development and 
statement is invoked by Sacks on behalf of a quite different intellectual and 
scientific enterprise, but the formal similarities in the problematics seem clear 
enough. 15 (And connections appear in other guises as well, for example, in the 

15 See the comments on the Fall 1965 lectures for further discussion of the 
relationship to generative grammar studies. 

In this regard as well, John Heritage has called to my attention an exchange 
involving Sacks and others at the Purdue Symposium on Ethnomethodology, in 
which he remarks in response to several inquiries (Hill and Crittendon, 1968: 41-2), 

One of the things that is obvious from the kind of analysis I have given you 
is that there can be a set of rules which can reproduce the problems in the data 
with which you started . . . [Query: How do you become satisfied with a 
solution?] ... I have a set of rules which give me back my data. 

xxii Introduction 

extensive paragraph numbering system which is used to organize ‘An initial 
investigation . . as well as the ‘Introduction’ of 1965 printed in this 
volume), a format hardly familiar to sociologists at all, but in common 
practice in linguistics at the time, though Sacks may well have come to it 
initially through his study of the law). 

I think it characteristic of Sacks’ relationship to work which he respected 
that it would enter into the warp and woof of his own thinking and would 
shape the way he did his work. And this is so not only in this formative stage 
of his work. Later on (in the work published in Schenkein (1978) but 
delivered as lectures 9-12 in Fall 1971), for example, his argument that the 
obscenity in a dirty joke is not its point, but is rather a form of ‘circulation 
control’ on knowledge which is packed or tucked in elsewhere, not overtly 
labelled or featured as the point of the joke, brings to bear a form of analysis 
developed by scholars of classical Greece such as Milman Parry (1971) and 
Eric Havelock (1963) in work on the role of the Homeric epics in an oral 
culture and its transformation in the passage from an oral to a literate culture. 

Another case in point is furnished by Cressey’s work on embezzlement 
(1953), which served Sacks (in ‘An initial investigation . . . ,’ 1972a) not to 
constitute the problem or suggest the shape of a solution, but as a way-station 
in the substantive analysis. Cressey had proposed as a key to understanding 
embezzlement that its perpetrators all had ‘a nonsharable problem.’ In Sacks’ 
effort to come to terms with the assertion by some avowedly suicidal persons 
that they had ‘no one to turn to,’ he proposed as a proximate solution that 
these persons found that what troubled them would, if recounted to the ones 
they would properly turn to (e.g., spouse), undermine the very relationship 
that made them ‘turnable-to;’ that is, precisely, they had a ‘nonsharable 
problem.’ But for Sacks this merely served to pose a problem: how to 
formulate the terms of the ‘search for help’ that yielded these persons as the 
candidates to be turned to, and therefore yielded the result that a problem not 
sharable with them left the searcher with ‘no one to turn to.’ And that 
recasting of the problem led to the central contribution of that analytic 
undertaking - the formulation of ‘membership categorization devices’ and 
their features. 16 

Sometimes Sacks would cite such sources. More often, the shape of the 
problem formation or solution, or the analytic resource, had simply entered 
into the currency of his thinking, and its source was lost sight of, especially in 
the context of lectures to undergraduate courses. The lecture format is, in this 
regard, ‘informal.’ Although published work which is, taken as a whole, 
remote from his concerns is often quoted directly and/or cited by name (e.g., 
Freud, Gluckman, Von Senden), more intimately related work is often not, 

16 In the paper presenting this work (Sacks, 1972a), the analytic ordering given 
in the text here is reversed. The paper begins with the most formal and general posing 
of the issues of categorization, and only eventually arrives at the more proximate, 
situated problem, as a ‘derivation,’ i.e., the dilemma presented when what qualifies 
another as the proper person to turn to will be compromised by the very turning to 

Introduction xxiii 

as for example (to cite an early instance in the text which follows) in the 
discussion of ‘common knowledge’ in lecture 3 of the 1964-5 lectures (as 
printed herein), for which Garfinkel clearly was relevant. In the preparation 
of these lectures for the present publication, that practice has not been 
addressed; it is a characteristic feature of the form in which Sacks’ work was 
shaped for presentation. 

As unexpected as may be the appearance in Sacks’ early lectures of echoes 
of the analytic style of transformational grammar, even more striking is the 
apparent lack of specific influences from the work of Erving Goffman. This is 
especially surprising since, during the years at Berkeley, Sacks took Goffman 
more seriously than he did virtually any other member of the faculty. 

At a very general level, of course, Goffman’s analytic enterprise had 
undertaken to establish the study of face-to-face interaction as a domain of 
inquiry in its own right, and his work was very likely central in recruiting 
Sacks’ attention to face-to-face interaction as a focus for the concern with 
practical theorizing and commonsense reasoning which animated the eth- 
nomethodological enterprise. Surely Sacks’ work, and work which it in- 
spired, have been important to whatever success and stability this area of 
inquiry has achieved. And Sacks could treat Goffman ’s work as setting a 
relevant domain for students for pedagogical purposes; in the first of the Fall 
1967 lectures, Sacks recommends readings in Goffman’s work as the most 
relevant sort of preparatory reading for the course, and the most indicative of 
the general stance of the course, while explicitly differentiating his own work 
from it. 

Goffman’s influence on Sacks was at its peak during Sacks’ years as a 
graduate student. While at Berkeley, for example, Sacks satisfied a require- 
ment in one of Goffman’s courses not with an empirical study of interaction 
of the sort chracteristic of his later work, but by writing the so-called ‘police 
paper’ (later published as ‘Notes on police assessment of moral character,’ 
1972c), concerned with methods of commonsense theorizing about appear- 
ances and moral character, and based on handbooks and manuals of police 
procedure. The subsequently published version of the paper begins with a 
handsome acknowledgement of debt to Goffman’s writing and lectures, and 
though the style and ‘address’ of the work differ in various respects from those 
of Goffman, the topic plays off of several themes recurrent in Goffman’s work 
at the time, and the exploitation of handbooks and manuals echoes Goffman’s 
use of manuals of etiquette and advice. But after this, Sacks’ work diverges 
increasingly from Goffman’s. 

To be sure, in later work Sacks addressed himself to more specific 
interactional topics mentioned in Goffman’s work (see, for example, the 
discussion of ‘rules of irrelevance’ in Goffman’s essay ‘Fun in games,’ (1961: 
19ff), or the passing mention of turn-taking (Goffman, 1964: 136), but the 
lines of influence are often not entirely clear. Goffman is reported to have 
responded to a question years later asking whether Sacks had been his student 
by saying, “What do you mean; I was his student!” Leaving aside the possible 
elements of generosity, irony and flipness in such a remark (and assuming 



that the report is, generally speaking, correct), a serious treatment of the 
directions of influence and the interplay of ideas between them remains to 
be written. 17 

That important divergences between Goffman and Sacks began to develop 
early on can hardly be doubted. These came to a head, both symbolic and 
practical, over Sacks’ PhD dissertation, an episode which cannot be recounted 
here. 18 For now the upshot must remain this: although in retrospect Sacks 
seems clearly to have labored in the same vineyard, and although he was not 
only formally Goffman’s student but learned a great deal from him, the 
degree to which Goffman influenced more specifically the work for which 
Sacks is known remains an open question. Certainly, such specific influences 
are not as much in evidence as most readers are likely to expect, either with 
respect to Goffman’s most characteristic substantive concerns - face, de- 
meanor, structures of attention and information, etc., with respect to 
governing themes - dramaturgic, ethologic, frame-analytic, etc., or with 
respect to data and method. 


In mentioning genres of work and particular people who constituted a 
relevant intellectual ambience for the early corpus of Sacks’ work, one name 
which might be thought missing is that of John Searle. But it turns out that 
Searle’s work constitutes a parallel stream, not a source. Indeed, although his 
Speech Acts was published in 1969, his paper ‘What is a speech act?’ appeared 
in 1965, the same year as the first of Sacks’ lectures. It is striking to compare 
the quite different tacks taken in these two approaches to the accomplishment 
of social action through the use of language, even if only in the brief and 
superficial way that space limitations compel. 

Searle begins not with a particular utterance - either actually spoken or 
invented. He addresses himself rather to a class of utterances that would 
satisfy whatever is required for them to effectively - felicitously - accomplish 
the speech act of ‘promising.’ It is the type ‘promises’ that provides Searle his 
object of inquiry. The solution takes the form of stating the “conditions . . . 

17 Some considerations on the relationship between Goffman’s work and 
conversation analysis may be found in Schegloff (1988). Goffman’s most explicit 
engagement with conversation-analytic work appeared in Forms of Talk (1981), the 
earliest of whose essays dates to 1974. 

The upshot was that Goffman found the argument of ‘An initial investiga- 
tion . . .’ circular, and no amount of discussion could move him from this view. Nor 
would he, for quite a while, step aside from the committee to allow its other members 
to act favorably on the dissertation, as they wished to do. Eventually, however, he 
agreed to do so, largely at the urging of Aaron Cicourel who, in the end, signed the 
dissertation as Chair of its sponsoring committee, making possible the awarding of 
the PhD in 1966. 



necessary and sufficient for the act of promising to have been performed in the 
utterance of a given sentence” (i.e., a general definition of ‘promise’), with a 
later derivation of the rules for performing acts of this class. 

Readers may recall the sort of result yielded by proceeding in this manner 
- the formulation of preparatory conditions, sincerity conditions, etc., 
followed by ‘‘rules for the function indicating device for promising.” The 
focus, then, is on the class or type of act, and the term describing it - 
‘promising.’ It is not on particular utterances or the contexts in which they 
occur. Indeed, Searle’s paper begins by invoking the most general context 
possible, ‘‘In a typical speech situation involving a speaker, a hearer, and an 
utterance by the speaker ...” 

Sacks’ first lecture starts in a significantly different way (and although the 
original transcripts show a much more uneven presentation than appears in 
the edited version, in the manner of their opening they do not differ). Sacks 
begins by offering particular utterances in a particular context. Our attention 
is focussed from the outset on particular exchanges, such as A: “Hello,” 
B: “Hello;” or A: “This is Mr Smith, may I help you;” B: “Yes this is 
Mr Brown;” or A: “This is Mr Smith, can I help you;” B\ “I can’t hear you,” 
which are 

. . . some first exchanges in telephone conversations collected at an 
emergency psychiatric hospital. They are occurring between persons 
who haven’t talked to each other before. One of them, A, is a staff 
member of this psychiatric hospital . . . 

Sacks goes on to offer a variety of detailed considerations about what these 
utterances, “This is Mr Smith,” “can I help you” or “I can’t hear you” might 
be observed to be doing, and how they might be doing it. Then he remarks 
(lecture 1, pp. 10-11): 

Clearly enough, things like “This is Mr Smith,” “May I help you”? and 
“I can’t hear you” are social objects. And if you begin to look at what 
they do, you can see that they, and things like them, provide the 
makings of activities. You assemble activities by using these things. And 
now when you, or I, or sociologists, watching people do things, engage 
in trying to find out what they do and how they do it, one fix which can 
be used is: Of the enormous range of activities that people do, all of 
them are done with something. Someone says “This is Mr Smith” and 
the other supplies his own name. Someone says “May I help you” and 
the other states his business. Someone says “Huh?” or “What did you 
say?” or “I can’t hear you,” and then the thing said before gets repeated. 
What we want then to find out is, can we first of all construct the objects 
that get used to make up ranges of activities, and then see how it is those 
objects do get used. 

Some of these objects { recall that ‘objects’ here refers to the utterances 
which have been examined } can be used for whole ranges of activities, 

xxvi Introduction 

where for different ones a variety of the properties of those objects will 
get employed. And we begin to see alternative properties of those 
objects. That’s one way we can go about beginning to collect the 
alternative methods that persons use in going about doing whatever 
they have to do. And we can see that these methods will be reproducible 
descriptions in the sense that any scientific description might be, such 
that the natural occurrences that we’re describing can yield abstract or 
general phenomena which need not rely on statistical observability for 
their abstractness or generality. 19 

Nor (one might add) do they rely for their abstractness or generality on being 
stripped of all contextual particulars (in the manner of Searle’s “In the typical 
speech situation . . .”) or on the stipulation of general constitutive definitions 
of verbs for speaking. 

The focus in Sacks’ work here, and in much of the work of the ensuing 
years, 20 is not on general constitutive conditions, or even on rules in Searle’s 
sense, but on practices and methods - on how Members, in particular contexts 
(or classes of context arrived at by examining particular contexts), methodi- 
cally construct their talk so as to produce a possible instance of an action or 
activity of some sort, and to provide for the possible occurrence next of 
various sorts of actions by others. 

Although the 1964-5 lectures exhibit some striking early explorations 
along these lines, a particularly exemplary instance of such an analysis is 
Sacks’ discussion in lecture 4 of Spring 1966, of the utterance by a previously 
present participant, after a newcomer to a group therapy session of teenaged 
boys has been greeted, “We were in an automobile discussion,” which Sacks 
undertakes to show to be “a possible invitation.” (In later ‘takes’ of this 
analysis, the treatment is varied; for example, in Fall 1968, lecture 6, (volume 
II) he discusses it as ‘orientational,’ although all the analysis bearing on its 
‘invitational’ aspect is included. This later discussion is rather fuller, more 
detailed and compelling.) 

His undertaking — . to build a method which will provide for some 

utterance as a recognizable invitation ...” - may sound like Searle’s, but it 
turns out to be quite different. There are two component tasks. One of these 
tasks is 

19 The reference to “reproducible descriptions in the sense that any scientific 
description might be” is an appearance in this first lecture of a theme and argument 
which Sacks had been percolating for some time, and which was written up at the end 
of the 1964-5 academic year in a putative introduction to a publication which never 
materialized. (That introduction is included in this volume, and its argument is 
recounted below, at pp. xxx-xxxii.) 

20 When Sacks does introduce a shift to a rather more general form of 
undertaking, for example at lecture 3 of the Fall 1968 set, it still has quite a different 
character than Searle’s. 



... to construct ... “a partial definition of an invitation.” What makes 
it partial is that while it’s a way of doing invitations, it’s not ... all the 
ways . . . there are other ways and those would be other partial 

The second task is to have this partial definition provide for the actual case 
which occasions the inquiry: 

We want to do both: Construct a partial definition of ‘invitation,’ and 
one that provides for ‘this is a case.’ 

It turns out that there are other things such an analysis should do, which need 
not preoccupy the present discussion. 

The construction of the method that provides for the data under 
examination as a possible instance of ‘invitation’ has two parts. First, Sacks 
characterizes the ‘slot’ in which this utterance occurs, and characterizes it in 
various ways - as (1) just after introductions and greetings, (2) in the arrival 
of a newcomer to a conversation already in progress, (3) in a situation of a 
psychiatric neophyte coming to group therapy for the first time and joining 
more experienced patients, etc. Second, he characterizes one particular aspect 
of the utterance itself - its formulation of the topic preceding the newcomer’s 
arrival as “an automobile discussion.” He shows that that formulation makes 
relevant the common category membership of the newcomer and the others, 
but a category membership as “teenaged boys” or potential “hotrodders,” 
rather than as “patients.” And in formulating the topic as one for which the 
newcomer might be competent in common with them (rather than as one for 
which he is not, as is done by a next speaker who extends the utterance by 
saying “. . . discussing the psychological motives for . . .”), a possible invita- 
tion is done. 

What this (here highly oversimplified) analysis provides, then, is not 
necessary and sufficient conditions for the felicitous performance of an 
invitation, or rules for its performance, but rather a partial method (Sacks 
refers to it as a “a partial definition”) for doing an invitation in a particular 
interactional / sequential context. 


As noted, the earliest lectures, of 1964-5, include a variety of efforts to 
develop analyses along these lines. Certain themes recur, only some of which 
can be remarked on here, to highlight something of an abbreviated catalogue 
of concerns animating Sacks’ work at the time. 

Consider, for example, the following sort of issue to which Sacks addresses 
himself recurrently throughout the 1964—5 lectures (this is not an exhaustive 

How to get someone’s name without asking for it (give yours), lecture 1. 

xxviii Introduction 

How to avoid giving your name without refusing to give it (initiate repair), 
lecture 1. 

How to avoid giving help without refusing it (treat the circumstance as a 
joke), lecture 2. 

How to get an account without asking for it (offer some member of a class 
and get a correction), lecture 3. 

How to get people to show they care about you, given few opportunities 
afforded by routine life, e.g., of the divorced (commit/attempt suicide), 
lecture 5. 

How to introduce a piece of information and test its acceptability without 
saying it, lecture 6. 

How to do a ‘safe’ compliment, i.e., without derogating others, lecture 8. 
How to get help for suicidalness without requesting it (ask ‘how does this 
organization work?’), lecture 10. 

How to talk in a therapy session without revealing yourself (joke), lecture 

12 . 

Sacks’ analytic strategy here is not a search for recipes, or rules, or 
definitions of types of actions. He begins by taking note of an interactional 
effect actually achieved in a singular, real episode of interaction (in the listing 
above, this often includes an achieved absence - something which did not 
happen). And he asks, was this outcome accomplished methodically. Can we 
describe it as the product of a method of conduct, a situated method of 
conduct, such that we can find other exercises or enactments of that method 
or practice, in that situation or context or in others, which will yield the 
accomplishment, the recognizable accomplishment (recognizable both to 
co-participants and to professional analysts) of the same outcome — the same 
recognizable action or activity or effect. 

So in the listing I have offered above, the ‘solutions’ mentioned in 
parentheses after some of the ‘problems’ are not ‘general;’ they are not 
practices which whenever or wherever enacted will yield those activities as 
systematic products. They are situated, contexted. How to describe the 
relevant contexts, the scope within which the proposed practice ‘works’? That, 
of course, is one of the prime sets of problems in this analytic enterprise. How 
shall we as analysts describe the terms in which participants analyze and 
understand, from moment to moment, the contexted character of their lives, 
their current and prospective circumstances, the present moment - how to do 
this when the very terms of that understanding can be transformed by a next 
bit of conduct by one of the participants (for example, a next action can recast 
what has preceded as ‘having been leading up to this’). Clearly enough, these 
questions are of a radically different character than those which are brought to 
prominence in an undertaking like that of Searle, or Austin (1962) before 

The recurrent theme documented above will remind some readers of 
‘indirect speech acts.’ In many items on that list the problem appears to be 
how to achieve some result without doing it ‘directly’ (as one says in the 



vernacular - and it is a vernacular term). The proposed ‘solutions’ might then 
be cast, in this vernacular and quasi-technical, idiom, as ‘indirect’ speech acts, 
although this is, of course, an idiom not employed within the conversation- 
analytic tradition, (cf. Levinson, 1983: 356-64 for one account). 

One line of inquiry (ibid., 274; Brown and Levinson, 1978; Lakoff, 1973) 
relates the use of indirect speech acts to considerations of politeness. But 
Sacks’ discussion focusses instead on what might be termed ‘strategic/ 
sequential’ considerations. He notes that the sorts of next turns made relevant 
by what might be called direct requests are quite different from the ones made 
relevant by the conduct whose methodic practices he is explicating. When 
answerers of the telephone at a psychiatric emergency service ask “What is 
your name?” they may get in return a request for an account - “Why?” - and 
may end up not getting the name. When they give their own names, they do 
not get asked “Why?,” because they have not done an action which is 
accountable in that way. The thrust of the analysis is, then, not considerations 
of politeness, but contingent courses of action as progressively and differen- 
tially realized in the set of turns that make up structured sequences based on 
what would later come to be called ‘adjacency pairs.’ 21 

The divergence of these two paths of analysis seems quite clearly related to 
the materials being addressed. On the one hand, we have single classes of 
utterances, and eventually (Searle, 1976) not even particular ones necessarily, 
but the categorical type of action which they are supposed to instantiate, 
singly and across contexts. On the other hand, we have particular utterances 
occurring in particular series of utterances, in particular organizational, 
interactional and sequential contexts, with the source of the utterance in prior 
talk and conduct accessible and demonstrably relevant to professional/ 
academic analysis as it was to the participants in situ and in vivo , and with the 
ensuing interactional trajectory which was engendered by the utterance 
inviting examination in the light of the set of possibilities from which it might 
have been selected. One of these sets of materials is the natural setting for the 
work of philosphy and ‘academic’ inquiry; the other is rather closer to the 
natural setting for the workings of talk in the everyday world. Sacks’ first 
lectures make clear what course is being set. 

The consequentiality of working with particular data, for example, with 
particular utterances, is underscored elsewhere in these and subsequent 
lectures, when Sacks directs the problematic of describing a ‘method for the 
production of . . . ’ to whatever action label one would assign to an utterance 
such as “I’m nothing.” Sacks asks (lecture 9, p. 67): how does someone 
“properly and reproducibly” come to say such a thing, this thing? What is 
someone doing by saying this thing, and how do they come to be doing it? 

At the time that Sacks was launching inquiry along these lines, a common 
reaction was that an utterance of this sort was ‘just a manner of speaking.’ 

21 Sacks deals with these themes from a different stance subsequently in Winter, 
1967, cf. the lecture for March 9 in particular, and the discussion of the varying tacks 
he takes below at pp. 1-li of this introduction. 



That the particular way of speaking, the phrasing, was almost accidental (a 
stance suitable to the view that an utterance is an enactment of a sentence 
which expresses a proposition, where it is the underlying proposition - 
perhaps accompanied by its ‘function indicating device’ - which finally 
matters, not the particulars which happen to give it expression on any given 
occasion). But Sacks saw it as the outcome of a procedure, as announcing ‘a 
finding’ by its speaker. He asked what that procedure was, and how it could 
arrive at such a finding, in a fashion that other participants would find 
understandable, and even ‘correct.’ He took seriously the particular form in 
which conduct appeared - the participants had said this thing , in this way , 
and not in some other way. He insisted on the possibility that that mattered 
- that every particular might matter. None could be dismissed a priori as 
merely (a word he particularly treated with suspicion) a way of talking. 

Of course, the fullest version of this sort of analytic undertaking was Sacks’ 
paper ‘An initial investigation . . .’ (1972a), where the utterance/action in 
question was ‘I have no one to turn to.” This utterance was also seen as 
reporting the result of a search, the description of which required developing 
the terms in which such a search might be understood to have been 
conducted, namely, ‘membership categorization devices.’ Early versions of 
parts of that paper (as well as other papers) can be found in the 1964-5 
lectures, for example in lecture 6. 

This way of working, then, mixed a kind of naturalism (in its insistence on 
noticing and crediting the potential seriousness of particulars of the natural 
occurrences of conduct) with the ethnomethodological concern for the 
Members’ methods for the production of a mundane world and commonsense 
understandings of it. Sacks asked how the recognizably detailed ordinary 
world of activities gets produced, and produced recognizably. It was just this 
way of proceeding - describing procedurally the production of courses of 
action - that Sacks understood at the time to be the foundation of the sciences 
as ‘science,’ and therefore the grounds for optimism about the principled 
possibility of a natural observational discipline in sociology. A brief account of 
this view (argued in the ‘Introduction’ by Sacks, Appendix I in this volume) 
is in order. 


Sacks had developed an argument 22 addressed to the question of whether 

22 The argument was written up, probably in the summer or autumn of 1965, 
after Sacks’ first academic year of lectures, as a possible introduction to a 
contemplated volume entitled The Search for Help. This publication, which was never 
pursued, would have included two papers - The search for help: no one to turn to’ 
(later published in Sudnow, 1972b), and ‘The search for help: the diagnosis of 
depression,’ never published. That the argument informed his thinking earlier, and 
entered into the first lectures, can be seen in the excerpt from lecture 1 cited at 
pp. xxv-xxvi above, and remarked on in n. 19. 



sociology could be shown to be a possibly ‘stable’ natural observational 
discipline. By this question Sacks meant to address the possibility that social 
science provided merely stopgap accounts of human action, conduct, behav- 
ior, organization, etc., until such disciplines as biology and neurophysiology 
matured to the point at which they could deal with such problems. (This was 
a position that Sacks was trying out when I first met him in 1961-2, and 
could be seen as a kind of riposte to Chomsky’s critique of Skinner. I always 
suspected that Sacks entertained the position as a provocation, in a law school 
pedagogical way, rather than as seriously tenable, but used it to force a 
consideration of the arguments necessary to set it aside. The position certainly 
shook me up when Sacks first confronted me with it in the winter and spring 
of 1962, for, in common with most sociology graduate students, I had treated 
such claims as long since undermined by Durkheim and other ancestors.) If 
sociology, or social science, were such a stopgap and thus ‘unstable,’ it hardly 
seemed worth investing much time and commitment in it. So before setting 
off on a serious research undertaking, it seemed in point to establish that a 
stable discipline was possible. Sacks believed that the argument he developed 
had a further pay-off; it showed something of the features the research 
enterprise and its results should have if it were to be, or contribute to, a stable 
science. The argument, briefly stated, was this. 

Contributions to science, including to sciences such as biology and 
neurophysiology, are composed of two essential parts. One is the account of 
the findings. The other is the account of the scientists’ actions by which the 
findings were obtained. What discriminates science from other epistemic 
undertakings is the claim that its findings are reproducible, and that 
reproducibility is itself grounded in the claim that the results were arrived at 
by courses of action reproducible by anyone in principle. Other investigators 
can, by engaging in the same actions, arrive at the same findings. 

Sacks argued that both of these parts of contributions to science are 
‘science’, and not just the findings. For it is the reproduction of the actions 
reproducing the results which make the findings ‘scientific’, and the descrip- 
tions of those courses of action which make their reproducibility possible. If 
the results are scientific, the descriptions of the actions for producing them 
must also be science. 

But, he noted, the descriptions of courses of action in scientific papers are 
not couched in neurophysiological terms, but take the form of accounts of 
methods or procedures. This form of account of action is reproducible, both 
in action and in description. 

So, Sacks concluded, from the fact of the existence of natural science there 
is evidence that it is possible to have (1) accounts of human courses of action, 
(2) which are not neurophysiological, biological, etc., (3) which are repro- 
ducible and hence scientifically adequate, (4) the latter two features amount- 
ing to the finding that they may be stable, and (5) a way (perhaps the way) 
to have such stable accounts of human behavior is by producing accounts of 
the methods and procedures for producing it. The grounding for the 
possibility of a stable social-scientific account of human behavior of a 

xxxii Introduction 

non-reductionist sort was at least as deep as the grounding of the natural 
sciences. Perhaps that is deep enough. 

This conclusion converges, of course, with the thrust of ethnomethodology 
as Garfinkel had been developing it, and was undoubtedly motivated, at least 
in part, by Sacks’ engagement with Garfinkel (and informed, perhaps, by 
Felix Kaufmann (1944) as well). Still, the argument is novel and provides a 
grounding from a different direction than Garfinkel had provided. For the 
tenor at least of Garfinkel’s arguments was anti-positivist and ‘anti-scientific’ 
in impulse, whereas Sacks sought to ground the undertaking in which he was 
engaging in the very fact of the existence of science. (And, indeed, in the 
earlier ‘Sociological description’ (Sacks, 1963) he had written, “I take it that 
at least some sociologists seek to make a science of the discipline; this is a 
concern I share, and it is only from the perspective of such a concern that the 
ensuing discussion seems appropriate.”) 


I have remarked on two types of problems taken up in the 1964-5 lectures 
- the reproducible methods by which ‘findings’ such as “I’m nothing” or “I 
have no one to turn to” may be arrived at (note in this regard the special claim 
on Sacks’ attention exerted by commonsense uses of ‘quantifiers,’ starting 
with the ones mentioned above,’ but extending to utterances such as ‘Everyone 
has to lie’, (Sacks, 1975)), and how to achieve some outcome without aiming 
for it ‘directly’. Several other recurrent themes in these earliest lectures might 
be mentioned here. 

One is an attention to certain ‘generic forms’ of statement or question, into 
which particular values can be plugged in particular circumstances. Sacks 
isolates, for example, the question form ‘Why do you want to do X?’ (lecture 
5, p. 33), or the generic form of statement ‘Because A did X, B did Y’ 
(lecture 5, p. 36). Later he focusses on the form, ‘X told me to call/do Y’ 
(lecture 10, pp. 76-7). It was very likely the exposure almost exclusively to 
calls to the Suicide Prevention Center, and the sort of recurrencies which they 
provided, which led to a focus on regularities so literally formulated. But it 
was in this sort of problem that the concern with the formats of utterances, 
often rather more abstractly and formally described, initially appeared. 

There is throughout these lectures the repeated use of ‘the socialization 
problem’ as a resource for focussing analysis. The question gets posed, ‘How 
does a child learn that X?,’ for example, that activities are observable; what 
properties of competence does socialization have to produce, and how are they 
produced; how does this learning take place (e.g., lecture 14, pp. 120-1). 
This form of problem or observation finds expression in Sacks’ writing of this 
period as well as in the lectures (for example, in the remarks in ‘An initial 
investigation. . . .’ concerning what is involved in learning how adequate 
reference is to be done), although it recedes in prominence in the later years 
of the lectures. 

Introduction xxxiii 

These early lectures of 1964—5 touch on, or give a first formulation of, a 
variety of themes more fully developed in later work, either of Sacks’ or by 

For example, although many believe that the early lectures were taken up 
with membership categorization, and that sequential organization is only 
addressed in later years, we have already seen that the early lectures - 
including the very first - engage that issue from the very beginning. To cite 
but one other instance of this early engagement, lecture 9 includes observa- 
tions on sequence organization (the asker of a question gets the right to do 
more talk), on what were later (Sacks et al., 1974) called contrasting speech 
exchange systems (remarks on press conferences and cross-examination), on 
how the turn-taking systems of different speech exchange systems can affect 
the forms of utterances (e.g., long questions when there is no right of follow 
up), and the like. 

Or note how the earlier-mentioned recurrent theme concerned with ‘how to 
do X without doing Y’ finds later resonance not only in Sacks’ work but in 
work such as that by Pomerantz (1980) on ‘telling my side as a fishing device’ 
(how to elicit information without asking for it), by Jefferson (1983) on 
‘embedded correction’ (how to induce adoption of a correct form without 
correcting the wrong one), and others. 

Or consider the material in lecture 1 1 concerned with glancing, looking, 
and seeing. The parts of this discussion which concerned the categories in 
terms of which one sees, anticipate the later discussion of ‘viewer’s maxims’ 
in the lectures on “The baby cried” (lectures 1 and 2 for Spring, 1966, 
eventually published as ‘On the analyzability of stories by children,’ 1972b). 
They display as well Sacks’ reflections on what such glance exchanges reveal 
about ‘norms’ in the more conventional sociological and anthropological 
sense, about ‘social integration,’ ‘alienation,’ and the like. And perhaps there 
is here as well a point of departure for Sudnow’s later (1972a) work on 
glances, for example in Sacks’ observation (p. 86) that “We start out with the 
fact that glances are actions.” 

It is worth noting that in some cases, discussions in these early lectures 
include points that are not found in later elaborations. Some of these seem to 
me to have been simply wrong - for example, the claim (lecture 5, p. 33) that 
‘opinion’ is something you don’t need a defense for. Others encountered 
problematic evidence within the conversation - analytic tradition of work. For 
example, Sacks had proposed that a method for doing greetings consisted in 
the use of one of the class of greeting terms in ‘first position.’ Schegloff (1967) 
disputes the generality of the claim by examining telephone conversation in 
which “Hello” in first turn is ordinarily not a greeting, and shows that claims 
in this domain of work can be addressed with data, investigated empirically 
and found to be the case or not. 

Still other portions of these early lectures, however, appear to be strong 
points which simply dropped out of later reworkings of the topic. For 
example, lecture 6 is a version of (or draws on) ‘An initial investigation . . . ,’ 
‘On the analyzability . . .’ ‘Everyone has to lie,’ and a paper which Sacks 



never published, entitled ‘A device basic to social interaction,’ concerned with 
the character of the categories which compose membership categorization 
devices as organizing devices for commonsense knowledge about members. 
But Sacks makes a point in this lecture which I do not believe ever appears 
in any of the other accounts of these domains, concerning the relativity of 
category collections such as age and class to the categorizer; he notes that the 
recipient of some utterance which includes some such categories (such as 
‘young man’) has to categorize the categorizer to know how they would 
categorize the one who had been categorized in the utterance. 

These lectures, then, have more than merely historical interest as embryonic 
versions of later developed work. Some of the themes here, however insightful 
and innovative, happen not to have been further developed. And others, 
which were further developed, left behind some points which are still valuable 
and can be found here. 


As with the 1964-5 lectures, those for the Fall 1965 term include first tries 
at topics (both accounts of specific data episodes and analytical topics raised 
from them) taken up and elaborated in subsequent terms, as well as 
discussions which do not get such subsequent development. Among the latter 
are, for example, ‘hotrodding as a test’ (lecture 10) or ‘non-translatable 
categories’ (lecture 12). Among the former are “The baby cried . . and 
membership categorization devices (lectures 1 and 2) more fully elaborated as 
lectures 1 and 2 in the following term, Spring 1966; collaborative utterances 
addressed via “We were in an automobile discussion” in Spring 1966 
(lectures 4 and 5); or ‘tying rules’ taken up in a number of subsequent lecture 

Still, there is good reason to read carefully the discussions of Fall 1965, 
even for topics which are given fuller, and apparently more satisfactory, 
treatment later. To cite but a single example, in the outline for the initial 
lecture on “The baby cried ...” (here appearing as Appendix A for Fall 
1965), at l.a.2 and a. 3, Sacks offers observations which do not appear in 
subsequent treatments of this material (either in Spring 1966, or in the 
subsequent publication as Sacks, 1972b) but which differentiate Sacks’ point 
here from other, parallel claims - often characterized as being concerned with 
the order of narrativity. Others (often more or less contemporaneously, e.g., 
Labov and Waletsky, 1966) have remarked that in narratives the ‘default’ 
organization is that order of sentences is isomorphic with the order of the 
occurrences which they report. And in later versions of this analysis Sacks 
seems to be making the same argument. As it is put in the published version 
(1972b: 330), “I take it we hear that as Sfentence] 2 follows Sentence} 1, so 
Occurrence] 2 follows Ofccurrence] 1.” But here, in the Fall 1965 outline, 
he notes that “this cannot be accounted for simply by the fact that SI precedes 
S2,” for “we can find elsewhere two sentences linked as these are, with 



nothing between, where we would not hear such an action sequence.” And he 
offers an instance from the same collection of children’s stories, . The 
piggie got hit by the choo-choo. He got a little hurt. He broke his neck. He 
broke his chin.” 

The point is that what is at work here is more than a matter of narrative 
technique or of discourse organization, although these may well be involved. 
Rather ‘commonsense knowledge’ of the world, of the culture, and of 
normative courses of action enter centrally into discriminating those actions or 
events whose description in successive sentences is to be understood as 
temporal succession from those which are not. It is not, then, a merely formal 
or discursive skill, but can turn on the particulars of what is being reported. 
This theme drops out of later discussions of these materials. 23 

If this point seems to resonate basic themes of so-called contextualist, or 
social constructionist or ethnomethodological stances, there are other elements 
in these early lectures which operate on a different wavelength. For example, 
early in the development of what he called ‘tying rules’ (in which he is 
addressing matters later often discussed under the rubric of ‘cohesion,’ cf., for 
example, Halliday and Hasan, 1976) he proposes (Fall 1965, lecture 5, 
p. 159) to be 

taking small parts of a thing and building out from them, because small 
parts can be identified and worked on without regard to the larger thing 
they’re part of. And they can work in a variety of larger parts than the 
one they happen to be working in. I don’t do that just as a matter of 
simplicity . . . the image I have is of this machinery, where you would 
have some standardized gadget that you can stick in here and there and 
that can work in a variety of different machines ... So these smaller 
components are first to be identified because they are components 
perhaps for lots of other tasks than the ones they’re used in. 

Thus, there is room within a larger, contextually sensitive, address to his 
materials (cf. the earlier-discussed contrast of Sacks’ starting point with that 
of Searle) for the recognition and more formal description of particular 
practices and sets of practices - here metaphorized as ‘gadgets’ or ‘machinery’ 
- which members can use in constituting coherent talk and specific lines of 
action and interaction, and for an appreciation that some of these may operate 
in a way substantially unqualified by the particulars of local context. 

Recall again (cf the discussion above at pp. xx-xxii) the echoes in Sacks’ 
work in this period of some of the themes of work in generative grammar 
(more accurately, an analytic model whose most lively embodiment at the 
time was generative grammar, but which is surely not limited to that domain 
of work). “ 4 The lectures for Fall 1965 were for a course whose catalogue title 

23 It does not drop out as a theme of the lectures, however; cf the discussion at 
pp. xxxvii-xxxviii below, and n. 26. 

2-4 It is worth making explicit here that Sacks kept himself informed of 



was ‘Culture and personality.’ Whether or not he would otherwise have been 
inclined to do so, it was perhaps this title which prompted some discussion by 
Sacks of the notion ‘culture.’ In setting out the orientation of his examination 
of the story told by a child, “The baby cried, the mommy picked it up,” Sacks 
subsumed it, and the ‘machinery’ by which it was produced and heard, under 
the notion ‘culture,’ of which he remarked, “A culture is an apparatus for 
generating recognizable actions; if the same procedures are used for generating 
as for detecting, that is perhaps as simple a solution to the problem of 
recognizability as is formulatable’ (Fall 1965, Appendix A, p. 226, emphasis 
in original.) 25 His description of what ‘the apparatus’ should do is strikingly 
reminiscent of lines from early Chomsky, and seems directly targetted at 
transformational grammar, but here, surprisingly, not at its principles - but 
at its product: ‘We are going to aim at building an apparatus which involves 
building constraints on what an adequate grammar will do, such that what an 
adequate grammar will do, some of the things it will do, we are going to rule 
out, and provide for the non-occurrence of’ (Fall 1965, Appendix A, 
p. 229). Sacks’ undertaking here seems in important ways to be shaped by 
the transformational grammar enterprise, albeit in a corrective stance toward 
it. The stance seems to be something like the following. Given an undertaking 
like the one generative grammar studies had seemed to set in motion, and 
operating with similar sorts of goals (e.g., to generate all and only the 
grammatical/acceptable sentences of a language), getting right results re- 
quires looking at something other than just the linguistic or, even more 
narrowly, the grammatical aspects of sentences or utterances. Not language, 
but culture, is the key object and resource. And while such an enterprise was 
understood by some ‘as ethnomethodology,’ by others it was seen as an 
anthropological/cultural version of cognitive science (albeit along different 

contemporary developments in a wide range of potentially relevant disciplines, and 
was aware of what seemed to be ‘hot’ topics and ways of working. His work 
recurrently speaks to such developments, sometimes explicitly, sometimes tacitly. He 
is aware of, and responsive to, his intellectual ambience. The present account often 
underscores such points of convergence and contrast - both with respect to the 
ambience at the time Sacks’ work was being done and with respect to developments 
at the time the present publication is being prepared. What may be of enduring 
interest is the larger picture of the intellectual stances and developments at issue, 
rather than the more transient excitements that pass over areas in ferment, even if 
these substantially engage a generation of workers in a field. 

25 On some readings, it is telling to compare this stance with Garfinkel’s account 
of ethnomethodology (1967: 1), about whose studies he writes, 

Their central recommendation is that the activities whereby members produce 
and manage settings of organized everyday affairs are identical with members’ 
procedures for making those settings ‘account-able’ . . . When I speak of 
accountable my interests are directed to such matters as the following. I mean 
observable-and-reportable, i.e., available to members as situated practices of 
looking-and-telling . 

Introduction xxxvii 

lines than those previously suggested by studies in ethnoscience and compo- 
nential analysis). 

There are various problem-types addressed and observations developed in 
these lectures which seem to have a (sociological?) bearing on what came to 
be called ‘cognitive science.’ Here I can mention only one of each. 

First, observations. Both in Fall 1965 (lecture 7) and in Spring 1966 
(lecture 18) Sacks comments on the differential ‘owning’ or control of certain 
categories by different social groups, and the not uncommon asymmetry 
between those to whom a category is applied and those who apply it. One 
particular focus for this line of analysis is the pair of terms ‘adolescent’ and 
‘hotrodder’ as applied to teenaged boys. ‘Adolescent’ is ‘owned’ by the 
conventional adult society, and is deployed by its members (together with all 
the commonsense knowledge or ‘conventional wisdom’ for which it is the 
organizational locus in the culture) more or less without regard to the views 
of those whom it is used to characterize. ‘Hotrodder’ (or, more recently, 
‘punker,’ etc.), on the other hand, are categories deployed by their incum- 
bents, and in ways often inaccessible to those who are not themselves 
members. It is this relative independence from the ‘official’ or conventional 
culture that led Sacks to term such categories ‘revolutionary’ (Spring 1966 
lecture 18, and Sacks, 1979). There seems to be here a whole area of inquiry 
which might be termed a sociology of cognition or a cognitive sociology quite 
distinct from other usages of this term (cf. especially Cicourel, 1974). Insofar 
as it involves the differential relevance of different category sets for the 
cognitive operations of persons dealing with categories of persons, its 
relevance to cognitive science seems transparent. 

Second, problem-types. There is a form of problem which Sacks takes up 
a number of times in the early lectures, each time on a distinct target, which 
can be best characterized as an ‘analysis of the ordering of cognitive 
operations’ (or the ordering of interpretive procedures). Two especially 
brilliant instances of solutions to this problem-type occur in the Spring 1966 
lectures. In lecture 1 1 (pp. 350-1) and again in lecture 21 (pp. 417-20), in 
dealing with an instance of ‘intentional misaddress,’ Sacks wonders how the 
co-participants in an interactional episode could have found who was being 
addressed, since the address term employed by the speaker (“mommy”) did 
not ‘actually’ apply to anyone present. He argues that, if they were finding 
‘who is being addressed’ by finding to whom the address term referred, then 
they would find no solution. Rather, he argues, they first use sequencing rules 
to find whom the current speaker would properly be addressing, and they use 
the product of that analysis in deciding how the address term is properly to 
be interpreted. He is thus able to sort out the order in which these analyses are 
conducted - first addressee, then address term - and it turns out to be just the 
opposite from what one might have thought. 

Another instance of the same problem is addressed in lecture 16 for Spring 
1966. Here the object of interest is what is conventionally known as ‘the 
possessive pronoun.’ Rather than taking a word like ‘my’ as indicating a 
relationship of ‘possessing’ toward whatever it is affiliated to (which yields 



results in usages such as ‘my brother’ or ‘my teacher’ which are either 
obviously faulty or in need of subsequent, and questionable, interpretation), 
Sacks argues that a hearer/receiver must first determine that what ‘my’ is 
attached to is a ‘possessable’ - the sort of thing which in that culture can be 
possessed (rather than a category from a membership categorization device, 
for example), in order to decide that ‘my’ is being used to claim possession. 
Once again, an ordering of analyses - of cognitive operations - seems clearly 

In both of these cases, the upshot of Sacks’ analysis is to reject as inadequate 
the view that linguistic items determine the meaning or the force of an action, 
and to insist instead that the cultural, sequential or interactional status of the 
objects employed in the utterance shape the interpretation of the linguistic 
item. 26 

But for Sacks there was no in-principle ordering of what sorts of things one 
consults first (e.g., the syntactic, semantic, sequential, interactional, etc.) and 
no necessary priority, therefore, among the disciplines which study them. 
Perhaps the first appearance of this problem-type is in lecture 4 for Fall 1965. 
Here Sacks is discussing various forms of ‘tying rules,’ forms of talk (such as 
indexical or anaphoric reference) which require a hearer to make reference to 
another utterance to understand a current utterance, and which thus ‘tie’ the 
utterances to one another. Encountering such usages of ‘that’ as “I decided 
that years ago” or “That’s the challenge,” Sacks remarks that they present a 
complication relative to other instances of tying procedures which he had 
previously discussed, for such usages must be distinguished from the use of 
‘that’ in, for example, “I still say though that if you take ...” Before 
analyzing a ‘that’ for the sequential tying connection it makes to some other 
(ordinarily prior) utterance, a hearer has to do a syntactic analysis to determine 
that the ‘that’ is the sort which can tie back to some earlier component of the 
talk. Here, once again, the sheer occurrence of an item (whether address term, 
‘my,’ or ‘that’) does not determine what is to be made of it. But whereas in 
the analyses previously discussed a linguistic analysis is contingent on prior 
sequential, interactional or cultural analyses, here the sequential ‘tying’ 
analysis is contingent on a prior syntactic one. 27 

26 Related discussions can be found throughout the lectures. For example, in the 
Spring 1966 lectures: lecture 11, pp. 350-1 (re how sequential and interactional 
organization controls the semantic and truth-conditional interpretation of an utter- 
ance, rather than the opposite, which is the ordinary understanding); lecture 16, 
p. 383; lecture 21, pp. 417-20; lecture 27, p. 451 (where sequential context is 
shown to control the very hearing of a word); and lecture 29, pp. 461-3. See also the 
earlier discussion at n. 23. Fuller discussion of the theme and the particular analyses 
on which it rests must await another occasion. 

27 Still, there is little doubt that the main thrust of analyses along these lines is that 
the understanding of talk is, in the first instance, controlled by the hearer’s grasp of 
the sequence in progress (or the sequential context more generally), rather than being 
derived from the linguistic tokens. Cf., for example, the discussion in Spring 1966, 
lecture 27, p. 451, where Sacks discusses the difficulty experienced by one participant 



Whatever the particulars, both these observations about control of 
categorization structures and deployments and the problem-type addressed to 
the ordering of cognitive or psycholinguistic or interpretive operations are 
theoretically central to the responsibilities of a sociological, or more gener- 
ally interactional, sector of what are now called the cognitive sciences. And 
to the degree that the results of these inquiries inform and constrain our 
understanding of how linguistic and category terms work, indeed can work, 
their import goes well beyond the interactional domain which is their initial 

The quasi-generativist themes in the Fall 1965 lectures, and in the 1964-5 
lectures as well, co-exist with analyses of particular action types (‘how to do 
action X’) based on empirical materials of talk, and co-exist as well with 
analyses of sequencing and tying practices - also developed on empirical 
materials, and addressed to the doing of conversation as an undertaking in its 
own right. This variety of topics and approaches (and I have not mentioned 
all the separate strands here) are, then, not a matter of stages in Sacks’ 
intellectual development over time. There are in these early lectures different 
sorts of undertaking underway, differentially developed by Sacks, differen- 
tially appealing to various segments of his professional readership, and 
perhaps differentially susceptable to development by others, and, therefore, 
differentially institutionalizable as a discipline. Surely, however, the drift of his 
own subsequent work favored some of these initiatives over others. 


If the lectures of Fall 1965 tilt in the direction of culture (whether incidentally 
because of the course title or because it was central to Sacks’ preoccupations 
at the time), the Spring 1966 lectures feature culture quite centrally. This was 
the most extensively taped and transcribed of the lecture sets, and it is as rich 
as anything in the materials assembled in these volumes. In its range - from 
the empirical detail of the interactional materials to discussions of some of the 
classic texts of social science and western culture - it gives the reader some 
sense of the power of the mind at work here, of the nuanced sensitivity to 
detail and of the scope of learning being brought to bear, and the distinctive 
stance being developed through the conjunction of these resources. Here I can 
touch only briefly on a few of the central themes of these lectures. 

One theme, clearly part of the ‘culturalist’ motif of these lectures, and 
surely not unrelated to the abiding preoccupation with ‘reflexivity’ and the 
‘incarnate character of accounts’ central to the continuing development of 
ethnomethodology in Garfinkel’s oeuvre, concerns the relationship between 

in hearing something addressed to him which is acoustically accessible to everyone 
else. He remarks that the party in question hears that turn by reference to the 
sequence in which it occurs ‘so as to hear, indeed, a puzzle, when he could hear 
something perfectly clear.’ 



‘commonsense knowledge’ and real world conduct or praxis on the one hand, 
and between commonsense knowledge and ‘professional’ inquiry on the other 
hand. This theme provides an opportunity as well to touch on the elements 
of continuity and discontinuity in the orientation of Sacks’ work going back 
to ‘Sociological description’ (1963). 

Although there is no direct connection between the positions explored in 
‘Sociological description’ and these lectures, there are echoes here, formal 
similarities to aspects of the earlier paper. By ‘no direct connection,’ I intend 
two observations. First, there is a substantial difference between what Sacks is 
doing in the lectures and the hypothesized program of studies which Sacks 
entertained in ‘Sociological description’ as a contrast with his depiction of 
how contemporary sociological inquiries are conceived and carried through. 
Second, there was no direct step-by-step theoretical development that led 
from the position taken up in the 1963 paper to the directions pursued in 
the lectures of 1964-6. On the other hand, I can only roughly suggest one 
sort of observation I have in mind in suggesting ‘echoes’ and ‘formal 

The central metaphor of ‘Sociological description’ was the so-called 
‘commentator machine,’ a ‘device’ describable (from one point of view) as 
composed of two parts - one which engages in some physical activity and 
another which produces a form of language, understandable as a description 
of what the first part is doing. Sacks entertains a variety of possible 
formulations of this device, and the ‘proper’ understanding of the relationship 
of its parts. The ‘doing part’ can be understood as a resource for coming to 
understand what the ‘speaking part’ is saying. The ‘speaking part’ can be 
understood as a description of what the ‘doing part’ is doing. The contraption 
may be understood as two independent devices. And so on. For those views 
in which the two parts do relate to one another, ‘discrepancies’ between the 
parts can be variously understood: for example, as the ‘speaking part’ offering 
inadequate descriptions of the ‘doing part;’ alternatively, as the ‘doing part’ 
malfunctioning and badly enacting the program set forth by the ‘speaking 

With such a theme in the background consider just a few elements of the 
first two lectures of Spring 1966 and some elements from the lectures of the 
intervening year, 1964-5. 

One of the central tasks which Sacks sets himself in the lectures on “The 
baby cried’’ is providing an account of how recognizable activities are done, 
and done recognizably. And in particular how the activity of ‘describing’ is 
done, and done recognizably. The key starting point here is that descriptions 
are recognizable, are recognizable descriptions, and are recognizable descrip- 
tions without juxtaposition to their putative objects. Much of Sacks’ effort in 
the early years of this analytic enterprise was given over to building an 
apparatus that provided recognizable descriptions without reference (by real 
life co-participants or by professional investigators) to what was putatively 
being described. The ‘membership categorization devices’ introduced in 
lectures 1 and 2 of Spring 1966, and the MIR device introduced in lecture 



6 of the 1964-5 lectures (p. 4l) 28 are key elements in such an apparatus. 
And the commonsense knowledge of the social world which is organized in 
terms of these categories, ‘protected’ as it is ‘against induction’ (as Sacks 
used to remark), provides for just such autonomously recognizable possible 
descriptions. When some potential discrepancy is suggested between what is 
provided for by the ‘knowledge’ organized around some category in a 
categorization device and what is observably the case about some putative 
incumbent of such a category, what may well be found (Sacks pointed out, 
and this is part of what he meant by ‘protected against induction’) is not 
the inadequacy of that ‘knowledge’ but rather the inadequacy of that 
person as a member of the category involved, an inadequacy which that 
person may feel and may seek to remedy. 

Although vastly transformed (from a ‘doing part’ and ‘speaking part’ to 
‘observable conduct’ and ‘recognizable description’), the problematics con- 
cerning (1) the proper juxtaposition of the practical activities of social conduct, 
(2) the commonsense knowledge of the mundane world and descriptive 
practices resident in that world, and (3) the proper formulation of investiga- 
tors’ stances and goals with respect to that world, persist from ‘Sociological 
description’ through these lectures. 

One component of these problematics is specially important throughout 
these lectures, surfacing at the end in Spring 1966, lecture 33 but also central 
at the beginning, and that is the relationship between commonsense knowl- 
edge which investigators may share with those whose conduct is the object of 
inquiry and the proper formulation of research questions, observations and 
findings. Sacks begins the discussion of “The baby cried’’ with a number of 
observations which he makes about the components of this little story, and 
offers the claim that his audience would have made (perhaps did in fact make) 
the same observations. But these are not sociological findings, he insists. They 
are simply the explication of commonsense or vernacular knowledge. Rather 
than constituting analysis, they serve to pose a research problem, namely, the 
construction of an apparatus that would generate (or that has generated) such 
observations, that would (in that sense) have produced them. And such an 
apparatus would constitute findings. 

Both parts of this analytic operation are important: making explicit the 
understandings which common sense provides of the world which members 
of the society encounter, including the conduct of others; and the provision of 
something that can account for those understandings. And it is important to 
keep them distinct and to insist on both. 

Consider, for example, the notion of category-bound activities. It is in order 
to address the observation that a report of ‘crying’ makes the category ‘baby’ 
(in the sense of a ‘stage-of-life’ category) relevant that Sacks introduces this 
notion, and the proposal that the activity ‘crying’ is ‘bound’ to the 

28 Sacks (ibid.) explains the term ‘MIR device’ by saying, “that is an acronym. 
‘M’ stands for membership, T stands for inference-rich, and ‘R’ stands for 



membership category ‘baby’ as one of the ‘stage-of-life’ categories in 
particular. But the observation that “crying is bound to ‘baby’’’ is (like the 
initial observations in the lecture) not a finding; it is merely the claimed 
explication of a bit of commonsense knowledge. As such it is just a claim, and 
cannot be simply asserted on the analyst’s authority. It has to be warranted 
somehow, either by a test of it or by requiring it to yield some further pay-off 
to analysis. 

And this is what Sacks does with “crying being category-bound to baby.” 
He immediately (lecture 1, p. 241) constructs a test of this category- 
boundedness, even though (as he says) “it’s obvious enough to you, you 
wouldn’t argue with the issue.” The pay-off, it will be recalled (lecture 1, 
ibid), is not only the explication of ‘praising/denigrating’ as a test for the 
category-boundedness of the action ‘crying,’ but an account for how to do 
such recognizable actions as ‘praising’ or ‘deprecating’, research goals already 
familiar from the 1964— 5 lectures and from elsewhere in the Spring 1966 
opening lectures. 

This stance is a basic and persistent one in these lectures. Elsewhere, for 
example, Sacks insists on testing the claim that the categorization device 
‘therapist/patient’ is ‘omni-relevant’ in the group therapy sessions which 
supply the data for most of these lectures (Spring 1966, lecture 6, p. 315; 
lecture 29, pp. 462-3; then again in Winter 1967, February 16, and Spring 
1967, lecture 14), although this claim can be treated as no less ‘obvious.’ To 
be sure, when he has recently made the point, Sacks sometimes asserts a 
claimed category-bound activity without carrying through a test or deriving 
a further finding (e.g., lecture 4, p. 302), but there can be little doubt that the 
principle is basic - commonsense knowledge cannot properly be invoked as 
itself providing an account, rather than providing the elements of something 
to be accounted for. 29 In my view, Sacks abandoned the use of ‘category- 
bound activities’ because of an incipient ‘promiscuous’ use of them, i.e., an 
unelaborated invocation of some vernacularly based assertion (i.e., that some 
activity was bound to some category) as an element of an account on the 
investigator’s authority, without deriving from it any analytic pay-off other 
than the claimed account for the data which motivated its introduction in the 
first place. 

The editorial effort to combine and blend largely overlapping treatments of 
the same material, which has prompted the inclusion of lectures delivered 
during the following term in Fall 1966, here in the Spring 1966 set (e.g., 
lecture 04. a), brings into relief certain shifts in analytic focus which 
accompanied a return by Sacks to the same empirical materials. Only two of 

29 See, for example, lecture 04. b, p. 295, here included with the Spring 1966 
lectures, though actually delivered later, in Fall 1966: ‘. . . it is our business to 
analyze how it is that something gets done, or how something is ‘a something,’ and 
not to employ it.’ 

This theme - as represented, for example, in the phrase introduced by Garfinkel, 
‘commonsense knowledge as topic and resource’ - is, of course, central to 



these shifts can be taken up here, and only for a brief mention. 

As remarked earlier, the analytic task set front and center in the initial 
lectures for Spring 1966 was “how recognizable actions get done and get 
done recognizably.” The first two lectures address those questions to the 
actions ‘doing describing’ and ‘doing storytelling.’ (The third lecture, omitted 
here because of its availability in a published version as ‘Everyone has to lie’ 
took up the issue of ‘doing a recognizably true statement.’) Lectures 04. a and 
04.b, interpolated here from Fall 1966, have a different analytic focus - 
observing and establishing orderliness - but lectures 4, 5 and 6 (delivered in 
the Spring term) continue the ‘recognizable actions’ theme (doing recogniz- 
able invitation and rejection) and reproducible methods for accomplishing 
recognizable actions. 

At the same time there is an apparent shift toward the invocation of a kind 
of evidence that was to assume an increasingly central place in Sacks’ 
conception of how to ground an argument or an observation. In lecture 4 
(from the Spring) he proposes that, in order to establish that “we were in an 
automobile discussion” is doing a recognizable invitation, it is necessary not 
only to agree that it “looks like an invitation” but to show “how that’s so” 
(p. 301) with the description of a method for doing invitations that works for 
the instance at hand. This echoes the stance of lectures 1 and 2. 

In lecture 04. a (pp. 286-7, 288-9) from the Fall 1966 term, Sacks offers 
as evidence that some earlier talk was attended by others than its overt 
interlocutors, and as evidence that it constituted a recognizable introduction, 
the prima facie evidence afforded by a subsequent speaker’s talk. Specifically, 
he notes, that when Ken responds to the utterance of his name by the 
therapist Dan not with “What” (as in an answer to a summons), indeed not 
with an utterance to the therapist at all, but with a greeting to the newly 
arrived Jim, he shows himself (to the others there assembled as well as to us, 
analytic overhearers) to have attended and analyzed the earlier talk, to have 
understood that an introduction sequence was being launched, and to be 
prepared to participate by initiating a greeting exchange in the slot in which 
it is he who is being introduced. 

There is a shift here in analytic stance and procedure, from the analyst’s 
understanding as initial point of departure on the one hand to the co- 
participant’s understanding as initial point of departure on the other. 

In the former mode, the analysis begins with an asserted convergence of 
interpretations and recognitions by the analyst and the analyst’s audience (for 
example, that something is a story, that ‘the mommy’ is ‘the mommy of the 
baby,’ that an utterance is doing an invitation, and so on). It proceeds by the 
provision of a methodical basis for both that convergence of understandings 
and the convergence between the ‘understanders’ and the producers of the 
to-be-understood ‘in the data.’ In the latter mode, analysis begins with an 
asserted observation (that not-overtly-engaged participants are attending, 
and, indeed, are obligated to attend to the talk), and then immediately 
grounds that observation in subsequent conduct by the co-participants in the 
episode being examined. That conduct is taken as displaying the product of 



their orientation to, and understanding of, the setting and what has been 
transpiring in it. The site of analysis is located in the setting of the data at the 
outset. And further: the analysts’ so treating the conduct of the participants 
is itself grounded in the claim that the co-participants so treat it. 

This contrast in stance and procedure is visible in this publication of the 
lectures only briefly, by virtue of the juxtaposition of the material from Spring 
and Fall 1966. What is seen only in lecture 0.4a-b here is seen increasingly 
thereafter, starting with the Winter 1967 lectures in the present volumes. Of 
course, this shift does not entail any abandonment of the commitment to 
provide an account for how the recognizable outcome - whatever sort of 
object it may be - is produced, although the form such an account might take 
does change over time. The subsequently developed description of the 
turn-taking organization, for example, is offered as a procedural account for 
how a substantial collection of observable achievements of ordinary talk are 
methodically produced by the co-participants. 

What I have referred to as the ‘culturalist’ tenor of the Spring 1966 
lectures is set in the first of its lectures, when Sacks sums up his initial gloss 
of the understanding of “The baby cried . . .’’as indicative of “the operation 
of the culture" as “something real and something finely powerful” (Lecture 
1(R), pp. 245-6, emphasis supplied). The analysis of the membership 
categorization device and of the commonsense knowledge organized by 
reference to its categories is, in its fashion, an analysis of culture - “an analysis 
of some culture,” as Sacks puts it (lecture 30, p. 469, emphasis supplied). 
Throughout these 34 lectures (cf. especially lectures 13, 16-21, 24-25 and 
3 1 and the appended manuscript ‘On some formal properties of children’s 
games’) may be found treatments of various forms and artifacts of ‘culture’ 
in at least that anthropological sense in which it refers to the categories 
through which ‘reality’ is grasped. Among these forms and artifacts are the 
categories of persons making up a society and its world and who is entitled 
authoritatively to ‘administer’ those categories (lecture 13), notions of 
possession and possessables, the constitution of observations and descriptions, 
measurements systems (lecture 24), games (lectures 13 and 31 and ‘On some 
formal properties . . .’), conceptions of danger and their bearing on differen- 
tially accomplishing such actions as warning and challenge (lecture 10, 12) 
etc. A kind of socio-cultural semantics is involved, and a largely anthropo- 
logical literature is invoked, reflecting Sacks’ engagement with then- 
contemporary work in so-called ‘ethnoscience.’ 30 

30 Cf. Sacks’ contrast of his own way of working on such matters with the 
then-mainstream approaches to ethnoscience, for example, with regard to ‘measure- 
ment systems,’ the discussion at lecture 24, p. 436, where the contrast is almost 
certainly with the work of Berlin and Kay (1969, but circulated in mimeo earlier) on 
color terms. 

Although ethnoscience is in point for this particular reference, Sacks’ reading in, 
and use of, the anthropological literature was very broad indeed - both in ‘areal’ 
terms and in ‘approaches.’ What he most appreciated was some combination of 
dense and acutely observed ethnography, tempered by a sharp theoretical intelli- 



All of these lectures provide rich materials for analysis and discussion, but 
in the present context, a brief consideration of Sacks’ treatment of games may 
serve to recall some of the relevant intellectual context for this sort of cultural 
analysis, as well as to permit a brief consideration of a direction for the study 
of culture and acculturation, including language acquisition, which deserves 
fuller exploration than it has been accorded. 

The most immediately relevant context for writing about games within 
American social science in the mid-1960’s traces back to the invention of 
‘game theory’ in 1944 by von Neumann and Morgenstern as a branch of 
mathematics with overtly ‘social’ applications (the title of their book was 
Theory of Games and Economic Behavior ), with its subsequent elaboration by 
economists and others concerned with strategic thinking, most visibly in the 
late 1950’s and early 1960’s, in authors such as Kenneth Boulding (1963), 
and Thomas Schelling (1961). The analytic force of the metaphor propelled 
it into the arena of discourse and interaction as well, the language of 
constitutive rules playing a central role in Searle’s development of speech act 
theory, for example, and strategic considerations entering psychology and 
sociology through varieties of ‘exchange theory’ (e.g., Thibaut and Kelley, 
1959, or Blau, 1964). More proximately to Sacks’ thinking, both Goffman 
and Garfinkel had explored the game model or metaphor in their own work 

- Goffman in his essay ‘Fun in games’ (in Goffman, 1961) and later in 
Strategic Interaction (1969, but written in 1966-7), and Garfinkel in the 
so-called ‘trust’ paper (Garfinkel, 1963), a paper from which he subsequently 
distanced himself, refusing to include it in the collection of his papers in 1967, 
Studies in Ethnomethodology. 

One problem with the assimilation of game theory into social science was 
in establishing the limits of its usefulness as a model of social reality, a concern 
surely central to both Goffman’s and Garfinkel’s treatment of it. One central 
objection is that ‘games’ fail as a basic model of social order much as ‘contract’ 
failed as a basic model in Durkheim’s discussion of ‘utilitarian’ social theory, 
an element of Durkheim (and Parsons’ (1937) treatment of Durkheim) 
especially emphasized by Garfinkel. In both cases, the ‘model’ - whether 
‘contract’ or ‘game’ — is itself ‘an institution,’ a normatively constrained 
organization of understandings and conduct, with its own constitutive 
infrastructure. ‘Contract’ could not undergird social order because, as a legal 
institution, it was itself undergirded by the social order it was invoked to 
explain. So also would ‘games’ fail as models of social interaction, for the 

gence, and informed by broad learning. I recall especially his appreciation of Hocart 
and Elizabeth Colson, of Fortune and Edmund Leach, of Evans-Pritchard and Max 
Gluckman. But less reknowned ethnographers were no less appreciated. His fondness 
for ethnography crossed disciplinary boundaries, and he collected original issues of 
the volumes produced by the founding ‘Chicago school’ of sociological field workers 

- Nels Anderson, Paul Cressey, Franklin Frazier, Clifford Shaw, Frederic Thrasher, 
Harvey Zorbaugh - and later sociological ethnographies such as Dollard (1937), 
Drake and Cayton (1945), and, in a different vein, studies like Cressey (1953), 
discussed earlier. 



conduct of games and their constitution presumed an infrastructure of 
interactional conduct, and an epistemic/ontological definition as a discrete 
order of ‘reality,’ within which games constituted a separate domain of 
activities. Such misgivings would surely have informed Sacks’ approach to 
games from the outset. 

It is noteworthy, in this regard, that Sacks focussed on childrens’ games, 31 
and that one of his central preoccupations was to get at that very infrastruc- 
ture by reference to which games, as a special class of events, also are 
undergirded. Thus, both in lecture 13 (on the game ‘Button-button who’s got 
the button’) and in the draft manuscript on children’s games appended to the 
Spring 1966 lectures, games are treated not as models of or about social life 
for the social scientist, but as training grounds for formal aspects of social life 
in social life, i.e., as arenas within social life for kids’ learning of central 
features of (the) culture, features such as the operation of membership 
categorization devices, the management of appearance and emotional display, 
etc. His treatment of children’s games aims to provide analytic particulars for 
his claim (‘On some formal properties . . . , Spring 1966, Appendix A, 
p. 502) that “Play then becomes an environment for learning and demon- 
strating criterial matters in real world action.” Games provide models of social 
life in social life for its initiates, and in that capacity can be looked to for 
methodically central components of culture. In that regard, for example, such 
a game-relevant contrast as ‘counting’ versus ‘not counting’ can provide 
materials on which can be built such ‘ real-world ’ contrasts as ‘legal versus 

Considerations of enculturation and ‘language acquisition’ provide an 
especially provocative focus for a matter which Sacks raises, in the first 
instance, rather more as a methodological point. Taking up the methodolog- 
ical relevance of sampling, Sacks points out that it depends on the sort of 
order one takes it that the social world exhibits. An alternative to the 
possibility that order manifests itself at an aggregate level and is statistical in 
character is what he terms the ‘order at all points’ view (lecture 33, p. 484). 
This view, rather like the ‘holographic’ model of information distribution, 
understands order not to be present only at aggregate levels and therefore 
subject to an overall differential distribution, but to be present in detail on a 
case by case, environment by environment basis. A culture is not then to be 
found only by aggregating all of its venues; it is substantially present in each 
of its venues. 

Leaving aside the consequences for the methodology of professional 
inquiry, consider the implication that . . . any Member encountering from 
his infancy a very small portion of it, and a random portion in a way (the 
parents he happens to have, the experiences he happens to have, the 
vocabulary that happens to be thrown at him in whatever sentences he 

3 1 Recall that this antedates by several years organized attention to play and games 
in the social science community, as represented, for example, in the wide ranging 
collection edited by Bruner, Jolly and Sylva (1976). 



happens to get) comes out in many ways pretty much like everybody else, and 
able to deal with pretty much anyone else’ (ibid., p. 485). 

In such a view, one might conjecture, we have one, and perhaps the major, 
theoretically available alternative to Chomsky’s argument that, given the 
highly limited and ‘degenerate’ sample of a language to which first language 
learners are exposed, most of language - the crucial part - must certainly be 
innate; they surely could not be induced from the available ‘inputs.’ 

The alternative is to consider a culture — and language as one component 
of culture - to be organized on the basis of ‘order at all points. ’ If culture were 
built that way, then socialization and language acquisition might well be 
designed accordingly, and require induction from just the ‘limited’ environ- 
ments to which the ‘inductee’ is exposed. As Sacks writes (ibid., p. 485), “. . . 
given that for a Member encountering a very limited environment, he has to 
be able to do that {i.e., grasp the order] . . . things are so arranged as to 
permit him to.” ‘Things’ here presumably includes the organization of 
culture, the organization of language, the organization of learning, and the 
organization of interaction through which the learning is largely done. What 
such a view projects is the need for an account of culture and interaction - and 
the acquisition of culture and language in interaction - which would 
complement a ‘cognitive’ language acquisition device and innate grammar 
much reduced from contemporary understanding. Studies relevant to such a 
view have been pursued for the last two decades or so, but not necessarily 
under the auspices of the theoretical stance toward culture which Sacks 
projects here. The evidence for an order at all points’ view has accrued 
throughout Sacks’ subsequent work and the work of others working in this 


The sessions from Winter 1967 appear in various respects transitional. There 
are returns to, and revisions of, themes initially discussed in earlier sets, 
including 1964-5 lectures, and initial explorations of topics taken up in much 
greater detail in subsequent terms. The discussion here can only touch on a 
few of these themes. 

It is in the session of March 2, 1967 that we find the first substantial 
consideration of turn-taking in multi-party settings. Here, as elsewhere in the 
lectures, a set of materials is treated lightly near the end of one term, and then 
is taken up in much greater detail in the next. The single session devoted to 
tum-taking in Winter 1967 is followed by seven lectures in Spring 1967 (the 
lectures on turn- taking from that term are not printed here), and an extensive 
run in Fall 1967. 32 

32 Another ‘take,’ embodying a different stance toward the work, is presented in 
the Fall 1968 lectures. 



A good deal of this treatment seems to have been prompted by reflections 
on the difference between the two-party talk discussed in the 1964-5 and Fall 
1965 lectures on the one hand (for which the materials were drawn from 
telephone calls to the Suicide Prevention Center), and, on the other, the group 
therapy sessions (GTS) on which many subsequent lecture sets are based. 

But the relevance of working with multi-person talk was not limited to the 
issue of turn-taking alone. To cite but one other product of the juxtaposition, 
the discussion at pp. 529-33 of the March 2 session is concerned with 
‘derivative actions,’ i.e., what a speaker may be doing to a third party by 
virtue of addressing a recipient in a certain way. This seemed to Sacks but one 
indication of the need to take up multi-person materials apart from two-party 
ones (p. 533). 33 

There is a theme taken up in the February 16 session, and touched on again 
on March 9 (pp- 543-6), whose relevance to contemporary concerns (both 
then and now) may be worth brief development here. One way of 
characterizing those concerns is the generic relevance of context to talk in 

The general question taken up is whether there is some way of formulating 
or invoking the sheer fact of the ‘settinged’-ness of some activity, without 
formulating or specifying the setting. The ‘solution’ which Sacks points to is 
the use of indicator terms (e.g., ‘here and now’ or stable uses of ‘this’) to do 
this, a usage which affords us evidence that it can, in fact, be done. Indicator 
terms can be seen as a machinery for invoking an unformulated setting, for 
referring to (categorially-) unidentifed persons, or taking note of unformu- 
lated activities. 

But where does this ‘question’ come from? Why is its solution of any 
interest? The beginning of the discussion, of course, is given not by a question, 
but by some observations which end up as the ‘solution.’ This was a common, 
and recommended, analytic procedure for Sacks: begin with some observa- 
tions, then find the problem for which those observations could serve as 
(elements of) the solution. 

The observations in point here concerned the use of such ‘indicator terms,’ 
terms whose special relevance for ethnomethodology had (under the name 
‘indexical expressions’) already been developed and underscored by Garfinkel 
(1967, passim). And the central observations here had come up in a train of 
considerations with a quite different focus, along the following lines. 

The discussion begins with the problems of the ‘professional’ analyst (i.e., 
the ‘conversation analyst,’ not the ‘therapist’ in the data) establishing the 
categorization device ‘patient/ therapist’ as omni-relevant for the participants 
(which would cast it as always-invocable - ‘on tap,’ so to speak - both by 
participants and by analyst). One way of doing that analytic task is to 
establish a formulation of the setting as ‘group therapy session’ as omni- 
relevant. Sacks then observes that this is but one form of ‘formulating as a 

33 Subsequently it turned out that derivative actions can be found in two-party 
conversation as well (Schegloff, 1984 [1976]). 



such-and-such,’ and that this is something that Members do. When they do it, 
it is consequential, that is, they are doing some possible action in doing the 
formulation. (Recall the discussion in Spring, 1966 of “we were in an 
automobile discussion’’ as a formulation of the topic as a such-and-such 
which is consequential - which does a possible invitation.) 

The question then is: is there some way of referring to the context, or 
components of the context, without formulating the context (or persons or 
actions in it) as such-and-such - without, therefore, potentially doing the 
actions which such a formulation might do. (Note that this can be a 
consideration both for members/participants-in-the-interaction and for pro- 
fessional analysts: for members so as to avoid doing the potential actions and 
the responses they would engender in the interactional setting; for analysts 
because it is precisely the escape from control by that interactional consequen- 
tiality, from what otherwise constrains or ‘disciplines’ formulations, that 
makes professional use of the lay device problematic). 

It is in this context that the observation about the indicator terms finds its 
resonance: terms like ‘here and now’ can invoke any present context and any 
conception of scope-of-context (‘in this room,’ ‘in 20th-century America,’ 
etc.) without formulating it. And by requiring a recipient to provide its sense, 
they recruit the recipient into the speaker’s project; they make the recipient 
complicit in forming up its sense. 

Several further brief comments will have to suffice here: 

1 The observation that formulating does more than simply naming what 
is formulated is focussed especially on ‘formulating what someone is doing’ in 
the March 9 lecture (pp. 544-6), and it sounds a theme central to the paper 
‘Formal properties of practical action’ (Garfinkel and Sacks, 1972), though 
that paper was (according to Sacks) Garfinkel’s work. In this discussion in 
Winter 1967 we see something of Sacks’ ‘take’ on similar issues, possibly one 
source of discussions of this theme between them. 

2 The considerations raised here (and when this theme is addressed 
elsewhere) impose a constraint on discussions of ‘context’ and its bearing on 
talk and action which has not been fully absorbed in the literature. The same 
problems raised about the categorization of persons/members pertain: 

the set of available characterizations is indefinitely extendable; 
the selection of some one or more is potentially a way of doing something, 
(i.e., is open to such understanding by others); 
in actual interaction, such possible interpretation by interlocutors and the 
responses they may offer in turn, can serve as a constraint on actually 
selecting such a formulation; 

the absence of such a constraint in the activities of professional analysts leaves 
the grounds of such choices undisciplined, and therefore problematic. 

The positivist solution to this problem (i.e., constraining the choice of 
formulation by explanatory adequacy as attested by ‘evidence,’ leaves the 



actual orientations of participants out of the picture. Where ‘context’ is made 
a central notion, these concerns will have continuing relevance. 

3 The effort to cast ‘therapist/patient’ as an omni-relevant categorization 
device has as one continuing relevance the concern with the bearing of gender 
(and perhaps other of what Hughes (1971) used to call ‘master statuses’) on 
talk-in-interaction. Were those who bring considerations of gender to bear on 
all phenomena of interaction to take seriously the considerations touched on 
just above, they might undertake to show that the pervasive relevance of 
gender can be grounded in the demonstrably equally pervasive orientations to 
it by participants to interaction. In effect, this would amount to showing that 
the categories and terms of gender identification are omni-relevant for 

Sacks’ exploration of this issue in Winter 1967 is left unresolved. By the 
time he comes to cast the indicator terms as ways of invoking the 
settinged-ness of the interaction without formulating it, the problem of 
establishing omni-relevance of either member-formulation or context- 
formulation has been abandoned. In its place is the possibility of non- 
formulation, of a kind of specific abstractness in treating the contexted 
character of activity. But the exploration of omni-relevance is taken up again 
in lecture 14 for Spring 1967 (cf. discussion below at pp. liii-livff.). 

One theme from the 1964-5 lectures which reappears in the Winter 1967 
lectures, reapplied to a related topic, is that of ‘direct’ versus ‘in-various- 
ways-non-direct’ speaking; the topic to which it is now applied is ‘euphe- 
mism’ (or what may be, once the data are examined, better termed ‘irony’). 
Although the 1964-5 lectures asked over and over again ‘how to do X 
without doing it overtly,’ the message here is that to ask why a euphemism 
or ironic trope was used instead of a direct or ‘literal’ saying is to get the 
question wrong. What Sacks is urging here (March 9, 1967, pp. 545-6) is 
that the first-order consideration is not directness/indirectness or literalness/ 
figurativeness. Rather it is (for the speaker) a saying which displays its 
relevance at that point in the talk, and (for the hearers) a saying such that 
their understanding (their capacity to understand) ‘proves’ the utterance’s 
relevance. The ‘norm’ is not, in the first instance, direct or literal reference, 
but rather ways of talking that are locally adapted and can show local 

The first-order considerations are thus tying rules and other local connec- 
tions between elements of the talk, rather than ‘saying it directly.’ The issue 
of ‘directness/indirectness’ comes to the fore only with academic analysts 
determined to understand the talk ‘in general,’ stripped of its local context. 
For them what comes to identify a bit of talk, to constitute its re-referable 
core, is its semantico-lexical content and perhaps its pragmatic upshot. With 
that as the core, then various ways of realizing that central identity can come 
to be formulated as more-or-less straightforward, direct, literal, or ‘tropic’ in 
some respect. What was in situ a production tailored to the details of local 
context is reinterpreted as a design for indirectness when local context is 

Introduction li 

stripped away and no longer accessible as the source of the utterance’s 

How is this line to be reconciled with the analysis in the 1964-5 lectures, 
where just this question is asked - e.g., why seek out the other’s name without 
asking for it directly? Perhaps this is one locus of development and change in 
Sacks’ thought during this period. But it is also possible that when the 
embodiment-of-indirection cannot be understood (by recipient, or by profes- 
sional analyst) as an adaptation to the local context, then the question of why 
the indirect rather than the direct may in fact be warranted and useful, and 
in just those terms. 34 


There are three predominants ‘casts’ to the lectures of Spring 1967. 

As noted earlier, the first seven lectures (not published in this edition) 
constituted the first sustained set on turn-taking, expanding the treatment in 
the lecture of March 2 in the Winter 1967 set. This is a ‘sequential 
organization’ cast. 

Lectures 8-9, earlier treated in lecture 3 for Spring 1966 and subsequently 
published as ‘Everyone has to lie’ (1975), have what might be termed more 
of a ‘socio-logic’ cast - juxtaposing to what might appear ‘logical’ ways of 
analyzing the conversational materials properly socio-logical ones. 

From lecture 1 1 on, the materials take on the same flavor of 
anthropological/cultural analysis that so heavily informs the Spring 1966 set. 
This is largely the result of a focus on membership categories underlying talk 
and relationships between those categories (their relative positionedness for 
instance), notions of activities ‘bound to’ those categories, and the sorts of 
commonsense ‘knowledge’ organized by reference to those categories (in the 
manner of ‘Y do X,’ where Y is a category name, such as ‘women,’ 
‘freshmen,’ ‘politicians,’ etc.) Some of this material was organized into a draft 
manuscript under the title, ‘On a device basic to social interaction,’ around the 
time of writing of ‘An initial investigation . . .’ As introduced into these 
lectures, much of the earlier statement seems to have been substantially 

The discussion here will be limited to some reflections on the “Everyone 
has to lie’’ analysis and on the reconsideration by Sacks of the matter of the 

34 And Sacks does sometimes work on an utterance by addressing, what it prima 
facie would be out of context, in a more-or-less ‘literal’ hearing, and with good 
results; cf. Spring 1966, lecture 29 pp. 461—2, where he shows how various 
components of the utterance “Usually there’s a broad in here” are neither produced 
nor grasped in their ‘bare’ literal sense: e.g., ‘here’, means not ‘this place’ but ‘when 
we are in {therapy] session;’ ‘a broad,’ means not ‘some woman’ but ‘the same 
woman,’ indeed a particular same woman,’ and one who is a member of the group, 



omni-relevance of formulations of setting and participants, earlier taken up in 
Spring 19 66 and in Winter 1967. 

At least one underlying source and rationale for the animating question 
being addressed in the ‘exercise’ concerned with the assertion “everyone has to 
lie” is formulated by Sacks (Spring 1967, lecture 8, p. 549) as “How could 
we as social scientists go about saying about something that a Member said, 
that it’s true.” It may be useful to ‘unpack’ the background for this question 
at least partially. 35 

As rhetoric as a core method and discipline for the analysis of what can be 
said gradually became demoted in the intellectual hierarchy of western 
culture, and logic developed an increasing hegemony, it brought with it an 
increasingly exclusive preoccupation with ‘truth’ as the paramount feature of 
assertions requiring definition and assessment. In part this concern was in the 
service of ‘science,’ and its aims of establishing stable propositions about the 
world whose truth could be established once and for all. 

When attention began in the 20th century to turn to statements in 
so-called ordinary language, the analytic apparatus available for use was that 
of formal logic, and it was in part by virtue of the results of applying a formal 
logic developed in the service of science and mathematics to ordinary language 
that natural languages were found defective and the need for ‘formal 
languages’ made compelling. But the goals of logic/science and ordinary 
discourse are by no means the same, and the use of language in them may be 
quite different. What is relevant to establishing the truth of a proposition in 
science - and what might be ‘meant’ by ‘truth’ - may be quite different from 
assessing the truth of a ‘commonsense assertion’ in ordinary circumstances. It 
is this gap which, in part, Sacks is addressing. 

Here, as elsewhere, Sacks’ exploration of this theme (the contrast between 
‘common sense’ and ‘scientific’ procedures) is focussed on a class of terms 
which is especially symbolic of logic — quantifiers. In ‘An initial investiga- 
tion . . .’as well such a term had become a focus of analysis. There it was the 
term ‘no one,’ in the claim by a suicidal person that they have “no one to turn 
to,” and Sacks undertook to explicate how ‘no one’ is used, and used 
‘correctly,’ given the ‘paradox’ that the assertion is made precisely in the 
conversation in which its speaker has turned to ‘someone.’ 

‘Initial investigation . . .’ showed how “no one to turn to” was not belied 
by having turned to someone for the conversation in which it was said because 
‘no one’ had as its scope only certain categories of person; ‘no one’ was not 
being used in some formal logical sense, as ‘no person.’ It was therefore 
misguided to begin with a ‘logical’ understanding of the term, when that was 
not the use being made of it in the production of the utterance. 

In lectures 8 and 9, the quantifier under examination is ‘everyone.’ Again, 
Sacks proposes not to begin with some sense of the term derived from logic 
(some ‘strict usage’ as he puts it), and find how trivially to disprove the 
assertion by showing that there is at least one person who does not have to lie. 

35 A similar question is taken up in Spring 1966, lecture 26. 



Rather, he proposes that we must investigate anew, and for its usage in 
ordinary conversation, how a term like ‘everyone’ is constituted and used. 

And more generally, assessing the truth of the assertion involves not just a 
manipulation of truth conditions, but rather an explication of those practices 
of talk-in-interaction which the assertion could reflect an orientation to, and 
whose actual operation could be what is being invoked in the asserted claim. 
In the context of this lecture, this refers to the contingencies of the ‘How are 
you’ question, its privileges of occurrence, its types of relevant answer, and 
how the further courses of action which its answers make contingently 
relevant affect the choice of answers in the first instance (pp. 556ff.). By the 
end of the discussion, this structure is generalized well beyond ‘How are you,’ 
and is used to specify where lying may be generically suspected, where 
confessions of it will be readily believed, etc. 

In any case, what emerges as criterial to the inquiry is not a logical analysis 
of the component terms of the assertion and an assesment of their combina- 
tion, but a social analysis of those contingencies of interaction which could 
give rise to the condition which the assertion claims. The upshot here is to 
blunt the prima facie application of ‘logical’ analysis as the first-order 
consideration in much the same fashion as several of the Spring 1966 lectures 
had the import of blunting the prima facie linguistic analysis of an utterance 
(cf. above at pp. xxxvii-xxxix, the discussion of ‘the ordering of analyses’). In 
both cases, the tools of linguistic and logical analysis are shown to have their 
relevance and applicability constrained by, and contingent on, prior sequen- 
tial, interactional and cultural specifications of the practices of talking 
underlying production of the utterance. 

What emerges is, then, a wholly different conception of what the analysis 
of ordinary discourse should consist in. It is this result which is adumbrated 
by asking at the outset how social scientists might go about assessing the truth 
of what a Member says, and this which animates that question. 3 

In lecture 14 (from p. 594 to the end of the lecture) Sacks again takes up 
the question of the ‘omni-relevance’ of a category collection. In the discussion 
of Winter 1967, the issue became redefined as invoking a context (and 
potentially associated membership categories) without actually formulating 
them - invoking the sheer fact of ‘settinged-ness’ (cf. above, pp. xlviii-1). 
Here, the discussion remains focused on the possibility of omni-relevance. 

What he means by ‘omni-relevance,’ Sacks says, is two-fold: ‘on the one 
hand, there are some actions which, for their effectiveness [i.e. , to be 
recognized as that type of action}, involve categorial membership in that 
collection, and, on the other hand, until the course of action is ended, one 
can’t rule out the further use of that collection.’ The elegant solution to the 
problem of showing ‘therapist/patient’ to be omni-relevant in the empirical 
materials under examination lies in noting that the effective doing of an 
ending to the occasion requires reference to the status of one of the parties as 

36 This sort of inquiry may be seen to inform the first paragraphs of lecture 1 1 as 



‘therapist.’ The point is made even more exquisite by ‘the therapist’ actually 
only hinting at the ‘session’s’ closure, and one of the more experienced 
patients interpreting that hint for a new patient. 

What is key to the solution is its focus on the efficacy of the utterance in 
implementing the action of initiating the ending of the session and the 
non-contingency of that action. Other actions could be understood to activate 
the relevance of the categories germane to their efficacy, but those categories 
might not on that account alone necessarily be claimable as omni-relevant. 
But accomplishing an ending is, first, a non-contingent occurrance for the 
occasion (the issue is not whether it will be done, but when), and therefore 
prospective, i.e., relevant even before an action might invoke it. It is this 
non-contingent prospective relevance of an action — an action which itself 
makes a membership category relevant - which grounds the argument for 
omni-relevance here. 37 

This lecture affords an especially clear example (as Sacks’ own lead-in 
makes clear) of one form which his kind of theorizing took. It regularly began 
with an observation about the particular materials being examined (an 
observation, of course, commonly informed by his prior work and wide 
reading). That observation might then be ‘developed:’ its terms being given 
an ‘anterior’ development, i.e., he would find and explicate what his own 
initiating observation could be seen, on reflection, to have presupposed; those 
presuppositions might well be more ‘observations,’ and more consequential 
ones. That package of observations might be followed up through discussion 
of matters in the literature which they touched off, through exploring purely 
formal kinds of logics they suggested, purely ‘theoretical’ possibilities they 
seem to entail, etc. But, recurrently, these ‘theoretical’ developments would be 
brought back to empirical materials - either what had initiated the whole line, 
or other materials which the line of theorizing brought to mind. It was in this 
sense that the effort was prosecuted to put theorizing at every point under the 
control of empirical materials. 

The actual presentations sometimes obscured this way of working. In 
lecture 14, for example, Sacks begins with what appear to be very abstract 
considerations about applying categories to partition a population, and the 
relationship between the partitionings yielded by different category collec- 
tions. This then is putatively ‘applied’ to the material at hand, in the analysis 
of ‘teenager/adult’ as a ‘cover’ collection preserving partitioning constancy 
with ‘patient/therapist;’ and in the covering of ‘patient/observer’ with 
‘performer/audience.’ It was initially an observation about the latter - re the 
utterance “Testing” (p. 593) in particular - which motivated much of this 
line. Of course, the most extensive such reversal of order of discovery and order 

37 How Sacks’ line of argument might bear on a claimed omni-relevance of gender 
(to re-pose an issue earlier discussed) is unclear. At the least, the constraint of “until 
the course of action is ended one can’t rule out the further use of that collection” 
requires working out in any occasion being examined, specifically what ‘the course of 
action’ can be taken to be. 



or presentation is the paper ‘An initial investigation . . . in which the 
originating observation was about “no one to turn to,’’ the serious exploration 
of which led to formulating it as the result of a search procedure, which 
required formulating the terms of the search and the categories by reference 
to which it is conducted, etc. It was with the last of these that the paper itself 


The Fall 1967 lectures turned out to be the last at UCLA. Sacks’ teaching 
during the Spring 1968 term was in seminar format, although he did offer 
sustained presentations on occasion, and these are included in the present 
volumes. And by Fall 1968 Sacks had moved to the University of California, 
Irvine (although there is no reason to think the prospect was already known 
at the time of the Fall 1967 lectures, or informed their delivery). 

These lectures include the first extended treatment of turn-taking presented 
in these volumes, although the first seven lectures for Spring 1967 (not 
printed here) represented Sacks’ actual first effort on this scale. The Fall 1967 
lecture set is the only one in which Sacks offered extended treatments of both 
turn-taking organization and tying structures. Tying structures are discussed 
in several earlier lecture sets, but not again after Fall 1967. And the 
discussions of identification and categorization to which Sacks returned several 
times in the lectures preceding Fall 1967 are not taken up here, and 
henceforth reappear only sporadically and for much briefer treatment. 
Sequential organization increasingly dominates the agenda of Sacks’ lectures, 
including expanding discussions of turn-taking, of sequence structure and 
adjacency pairs, of overall structural organization, of story-telling organiza- 
tion, etc. 

If the Spring 1966 lectures were especially ‘anthropological’ in orientation, 
then the Fall 1967 lectures are especially oriented to linguistics. 

This note is sounded early, when in the initial lecture, a general introduc- 
tion, Sacks (pp. 622-3) projects the preoccupations of the course with 
‘sequential analysis’ (though not under that name), which he introduces by 
remarking that ‘ . . . the discoverable aspects of single utterances turn out to 
be handleable - perhaps handleable only - by reference to sequencing 
considerations . . . ,’ and declaring his interest in “. . . how it is that sequenc- 
ing considerations turn out to be implicative of what happens in a given 

“Linguistics,” by contrast (he argues), “is that study of the utterance which 
involves detecting those features of it which are handleable without reference 
to such considerations as sequencing; i.e., without reference to that it has 
occurred in conversation” (ibid.). 

One question, then, is whether “there is the possibility of. . .a fully 
comprehensive, coherent linguistics without such matters.” Another is how 
such study of single utterances can be “brought into alignment with what we 



know about sociology and anthropology. And if not, what then?” 38 

Recurrently throughout these lectures Sacks brings the results of a line of 
analysis or argument into juxtaposition with the main thrust of contemporary 
linguistic theory and analysis (i.e. , of the early to mid-1960s). One result is 
the sketching of whole orders of observable regularity and apparent normative 
organization which have largely, in some cases entirely, escaped the notice of 
the main thrust of the contemporary study of ‘language.’ In some respects, 
this is undoubtedly related to the ambition of modern linguistics (tracable at 
least to de Saussure) to transcend particular contexts and media of language 
use - not only social and cultural settings, but also oral and written 
embodiments - so as to describe an underlying, presumably invariant, 
linguistic code. The attention to sequential organization - an order of 
organization seemingly inescapable in the effort to understand and describe 
actual, naturally occurring talk in interaction - forcefully belies the premise of 
the currently dominant commitments of linguistics. Running through both 
the Fall 1967 lectures and the presentations of Spring 1968 are several 
recurrent themes, whose central upshot is: 

How sequential considerations necessarily inform or bear on the construction 
and understanding of single utterances; 

How understanding of some talk is regularly displayed by its recipients; and 
What that has required of recipients, and how those requirements are 
formative of their talk in turn. 

These themes are returned to persistently, almost compulsively, and they are 
considerations of a ‘foundationalist’ sort - that is, they go to the matter of 
what foundations a discipline of language must be understood to rest on. 

Sacks has seemed to some to have abandoned his commitment to 
contextually-sensitive analysis in turning to the study of sequential structure, 
and turn-taking in particular. But in insisting on the decisive relevance of 
sequential organization as furnishing the most proximate reference points of 
context, Sacks showed the consequences of disattending the fact that language 
was being used in a medium which was inexorably temporal and interac- 
tional. The results of these explorations of sequential context offer, in their 
own way, as sharp a contrast to formal linguistic analysis as did Sacks’ earlier 
explorations in the 1964-5 lectures offer a contrast with Searle’s efforts at 
context-free speech act theorizing (cf. above, pp. xxiv-xxix). 

38 Later (for example, in a letter to me in 1974) Sacks seems to have taken a 
different tack, namely, that a systematic discipline might not be buildable on the 
analysis of single utterances, or single instances of other units or occurrences, but that 
large amounts of material might be needed. At the time of his death, we had just 
begun a large-scale investigation of ‘next turn repair initiators’ which was going to be 
an exploration of that sort of undertaking. This subsequent development, of course, 
in no way blunts the impact which Sacks produced by asking what was to be made 
of the single utterance or the single sequence or the single exemplar of anything to be 
analyzed, and the detailed findings which this way of working led him to. 



Appreciation of the recurrent linguistic orientation in many of these 
lectures should not be allowed to obscure the range and variety of matters 
taken up in them, and the diversity of the intellectual resources being called 
upon from many different traditions of inquiry. One case in point must 

In lecture 6 for Fall 1967 Sacks returns to a point which had come up in 
earlier sets of lectures, concerning the inclusion in analysis of things which did 
not happen, here offering as one special relevance of ‘next-speaker selection 
techniques’ and ‘paired utterances’ (the later ‘adjacency pairs’) that they 
provide enhanced analytic leverage for speaking of something being absent - 
e.g., the utterance of an unresponsive selected next speaker, or the absence of 
a responsive paired utterance. The problem of warranting claims about 
‘absences’ has resonated to many corners of the conversation-analytic domain 
of issues. Then Sacks adds (p. 670): 

A way, perhaps, to develop a notion of ‘absence’ involves looking to 
places where such a notion is used and attempting to see whether there 
are various sorts of relevance structures that provide that something 
should occur. Parenthetically, I’ll give as a rule for reading academic 
literature, that whenever you see somebody proposing that something 
didn’t happen - and you’ll regularly find, e.g., sociologists, anthropol- 
ogists, or historians particularly, saying that something didn’t happen, 
something hadn’t been developed yet - that they’re proposing that it’s 
not just an observation, but an observation which has some basis of 
relevance for it. 

Sacks’ interest in the matter of ‘absences’ antedates his work with conversa- 
tional materials. He had taken a special interest in an observation of Max 
Weber’s that some aspect of ancient Middle Eastern history was to be 
understood by reference to the fact that (as Sacks would put it in conversation) 
“that was before the appearance of the horse as an instrument of warfare.” 39 
The issue this posed was, how could something be the consequence of 
something which had not happened yet? Clearly some set of relevancies to 
which the theorist was oriented informed this way of thinking. 

And, earlier yet, I recall a conversation at the Law and Society Center in 
Berkeley in 1962-3 (involving Sacks, a Marxist graduate student in sociology 
from Argentina and myself) in which the discussion lingered on ‘explanations’ 
for the absence of revolutions founded on the Marxist notion of ‘false 
consciousness.’ At issue were both the theoretical status of observations 

39 Weber (1952: 6, emphasis supplied): 

Because the nature of military and administrative technology of the time 
precluded it, before the seventeenth century BC, a lasting political conquest 
was impossible for either of the great cultural centers. The horse, for instance, 
while not completely absent, at least, not in Mesopotamia, had not as yet been 
converted into an implement of special military technique. 



concerning the non-occurrence of revolution, and the reliance, in the concept 
of ‘false consciousness,’ on a stipulated account by the theorist/analyst of 
what the ‘real’ interests of the proletariat were, a correct appreciation of which 
was ‘absent’ from their (i.e. , workers’) understanding of the world. What 
made those ‘understandings’ relevant, such that not sharing them amounted 
to their ‘absence,’ and rendered other beliefs of the working class to be ‘false 
consciousness,’ with sufficient explanatory power to account for the absence of 

So when Sacks refers in lecture 6 to a ‘rule for reading academic literature,’ 
there is specific background informing the line he is recommending. Having 
initially engaged this issue in the social science literature, Sacks came to find 
it illuminated in his engagement with interactional materials. For the 
underlying ‘logic’ was, although encountered in the first instance in academic 
materials, but an aspect of ‘commonsense’ or ‘practical’ theorizing which had 
been incorporated in professional social science theorizing. 

Eventually Sacks pursued this matter with a variety of interactional 
materials. For example, in one of the 1964-5 lectures he remarks on the 
special intimacy and power of a line reportedly addressed to a beloved in 
explanation of some past bit of biography, “That was before I met you, and 
I was lonely then.’’ Here again a ‘state-of-the-world’ is explained by 
something that had not yet happened, in a powerful display of retroactive 

So these lectures of Fall 1967, however oriented to exploring their interface 
with contemporary linguistics, retain their grounding in social (even ‘socio- 
logical’) and cultural analysis. Indeed, it is at the meeting point of these 
disciplines that the analytic action of these lectures is situated. 


This volume presents roughly the first half of those lectures which Sacks chose 
to tape record and have transcribed. The introduction to this point has 
attempted to provide some thematic overview of these lectures, and some- 
what more detailed background and exploration of a few selected issues. 

This effort at an overview has been truly daunting, indeed, beyond my own 
capacities at the present time. Part of this may surely be traced to my own 
shortcomings. But, for the most part, it reflects rather the extraordinary 
richness and multi-facetedness of Sacks’ corpus. In its variety, depth, and 
freshness of vision it defies domestication into convenient guidelines to a 
reader. At least part of this derives from the methodological character of 
Sacks’ initiative - the new way of working he introduced. Starting out with 
a commitment to lay bare the methodicity of ordinary activities, and with his 
talent for seeing in singular occurrences the structural elements of which they 
were formed and composed, a world of data which refreshed itself every 
moment more than a legion of Sackses could ever make a dent in provided 
a virtual infinity of opportunities for new observations, and new orders of 



Not that it was easy! Sacks often complained about how hard the work 
was, and that it did not seem to get easier. He spoke in the early 70s of giving 
it up and working on something less demanding. The problem was, he 
observed, the need to see “around the corner,” to penetrate through the 
blinders of the implacable familiarity of the mundane materials with which 
we worked, and the commonsense models and expectations derived from a 
social science which had never addressed itself to the simple observational 
tasks of a naturalistic discipline in which such models ought to have been 
grounded in the first instance. If we were to try to build a discipline, we 
needed to be able to be freshly open to what could be going on in any given 
piece of interaction, and to how activities and conduct could possibly be 
organized. And it was hard to say which was more difficult - to see clearly 
what was going on in some bit of material, or to figure out how to build from 
such observations and analyses a worthy discipline. And, of course, these were 
not independent orders of task - for how to address the empirical materials 
was always being informed by the direction in which it appeared a discipline 
might be pursued, and one surely wanted the character of the discipline to be 
shaped centrally by one’s sense of how social activities were actually 

In any case, the main line of engagement for Sacks was in directly taking 
up particular occurrences, particular bits of tape and transcript. And in 
leaving as open as he could what there was to be noticed about that bit of 
occurrence, what there was to be learned from it, what we might get to see the 
importance of for the first time. And this insistence on freeing each next 
engagement with data from the past - not only the past of the social sciences, 
but also past work of this sort, including (especially) his own — while still 
allowing it somehow to inform analysis is what allowed each new fragment of 
data, each next look back at an old fragment of data, to provide a possible 
occasion of discovery. Although the sorts of things which emerged (however 
rich and multifaceted) were constrained by the particular metier of his mind, 
their range was truly astounding. They overflow efforts to contain them and 
package them for overview. 


Anderson, Alan Ross and Moore, Omar Khayyam, I960. Autotelic folk models. 
Sociological Quarterly, 1, 4, 203-16. 

Atkinson, J. Maxwell and Heritage, John C. (eds), 1984. Structures of Social Action 
(Cambridge: Cambridge University Press). 

Austin, John L., 1962. How to Do Things With Words (Oxford: The Clarendon Press). 
Barker, Roger G. and Wright, Herbert F., 1951. One Boy’s Day (New York: Harper and 

Barker, Roger G. and Wright, Herbert F., 1954. Midwest and its Children (Evanston, 
III.: Row Peterson and Co.). 

Berlin, Brent and Kay, Paul, 1969. Basic Color Terms: their universality and evolution 
(Berkeley and Los Angeles: University of California Press). Previously circulated as a 

lx Introduction 

Working Paper by the Laboratory on Language Behavior Research, University of 
California, Berkeley). 

Blau, Peter, 1964. Exchange and Power in Social Life (New York: John Wiley). 

Blumer, Herbert, 1939. An Appraisal of Thomas and Znaniecki’s The Polish Peasant in 
Europe and America (New York: Social Science Research Council). 

Boulding, Kenneth, 1962. Conflict and Defense: a general theory (New York: Harper). 

Brown, Penny and Levinson, Stephen C., 1978. Universals in language usage: politeness 
phenomena. In Esther Goody (ed.), Questions and Politeness: strategies in social 
interaction (Cambridge: Cambridge University Press), 56-310. 

Bruner, Jerome S., Jolly, Alison, and Sylva, Kathy, (eds), 1976. Play: its role in 
development and evolution (New York: Basic Books). 

Button, Graham and Lee, John R. E. (eds), 1987. Talk and Social Organization 
(Clevedon: Multilingual Matters). 

Chomsky, Noam, 1957. Syntactic Structures (The Hague: Mouton). 

Chomsky, Noam, 1965. Aspects of the Theory of Syntax (Cambridge: MIT Press). 

Cicourel, Aaron, 1974. Cognitive Sociology : language and meaning in social interaction 
(New York: Free Press/Macmillan). 

Cressey, Donald, 1953. Other People's Money (Glencoe, III.: The Free Press). 

Dollard, John, 1937. Caste and Class in a Southern Town (New Haven: Yale University 

Drake, St. Clair and Cayton, Horace, 1945. Black Metropolis (New York: Harcourt 
Brace & Co.). 

Garfinkel, Harold, 1959. Aspects of the problem of commonsense knowledge of social 
structure. Transactions of the Fourth World Congress of Sociology, 4, 51-65. 

Garfinkel, Harold, I960. The rational properties of scientific and commonsense activities. 
Behavioral Science, 5, 72-83. 

Garfinkel, Harold, 1963. A conception of, and experiments with, ‘trust’ as a condition 
of stable concerted actions. In O. J. Harvey (ed.). Motivation and Social Interaction 
(New York: Ronald Press), 187-238. 

Garfinkel, Harold, 1964. Studies of the routine grounds of everyday activities. Social 
Problems, 11, 225-50. 

Garfinkel, Harold, 1967. Studies in Ethnomethodology (Englewood Cliffs, NJ: Prentice 

Garfinkel, Harold and Sacks, Harvey, 1972. On formal structures of practical actions. In 
John C. McKinney and Edward A. Tiryakian (eds), Theoretical Sociology (New York: 
Appleton-Century-Crofts) . 

Goffman, Erving, 1959. The Presentation of Self in Everyday Life (New York: Doubleday 

Goffman, Erving, 1961. Encounters: two studies in the sociology of interaction (Indiana- 
polis: Bobbs-Merrill). 

Goffman, Erving, 1963. Behavior in Public Places (New York: The Free Press). 

Goffman, Erving, 1964. The neglected situation. In John J. Gumperz and Dell Hymes 
(eds). The Ethnography of Communication , special issue of the American Anthropologist, 
66, 6, part 2, 133-6. 

Goffman, Erving, 1969. Strategic Interaction (Philadelphia: University of Pennsylvania 

Goffman, Erving, 1971. Relations in Public (New York: Basic Books). 

Goffman, Erving, 1981. Forms of Talk (Philadelphia: University of Pennsylvania Press). 

Halliday, Michael A. K., and Hasan, Ruqaia, 1976. Cohesion in Discourse (London: 

Havelock, Eric A., 1963. Preface to Plato (Cambridge, Mass.: Harvard University Press). 

Hill, Richard J, and Crittenden, Kathleen Stones, (eds), 1968. Proceedings of the Purdue 
Symposium on Ethnomethodology, Institute Monograph Series No. 1 (Institute for the 



Study of Social Change, Department of Sociology, Purdue University). 

Hughes, Everett C., 1971. Dilemmas and contradictions of status. In The Sociological Eye: 
selected papers (Chicago and New York: Aldine and Atherton), 141-50. Reprinted 
from The American Journal of Sociology, 50 (1945). 

Jefferson, Gail, 1983. On embedded and exposed correction in conversation. In Studium 
Linguistik, 14, 56-68. Reprinted in Graham Button and John R. E. Lee (eds). Talk 
and Social Organization (Clevedon: Multilingual Matters), 86-100. 

Jefferson, Gail, 1989. Editor’s notes. Human Studies, 12, 395-404. 

Kaufmann, Felix, 1944. The Methodology of the Social Sciences (New York: Humanities 

Kneale, William and Kneale, Martha, 1962. The Development of Logic (Oxford: The 
Clarendon Press). 

Labov, William and Waletsky, Joshua, 1966. Narrative analysis: oral versions of 
personal experience. In J. Helm (ed.), Essays on the Verbal and Visual Arts (Seattle: 
University of Washington Press), 12-44. 

Lakoff, Robin, 1973. The logic of politeness: or minding your p's and q’s. Proceedings of 
the Ninth Regional Meeting of the Chicago Linguistic Society, 292-305. 

Levinson, Stephen C., 1983. Pragmatics (Cambridge: Cambridge University Press). 

Opie, Iona and Opie, Peter, 1959. The Lore and Language of Schoolchildren (Oxford: The 
Clarendon Press). 

Parry, Milman, 1971. The Making of Homeric Verse: the collected papers of Milman Parry, 
edited by Adam Parry. (Oxford: The Clarendon Press). 

Parsons, Talcott, 1937. The Structure of Social Action (Glencoe, Illinois: The Free Press). 

Pomerantz, Anita, 1980. Telling my side: ‘limited access’ as a ‘fishing’ device. Sociological 
Inquiry, 50, 3-4, 186-98. 

Psathas, George (ed.), 1979. Everyday Language: studies in ethnomethodology (New 
York: Irvington Press). 

Sacks, Harvey, 1963. ‘Sociological description.’ Berkeley Journal of Sociology, 8, 1-16. 

Sacks, Harvey, 1972a. An initial investigation of the usability of conversational data for 
doing sociology. In David N. Sudnow (ed.), Studies in Social Interaction (New York: 
The Free Press), 31-74. 

Sacks, Harvey, 1972b. On the analyzability of stories by children. In John J. Gumperz 
and Dell Hymes (eds), Directions in Sociolinguistics (New York: Holt, Rinehart and 
Winston), 325-45. 

Sacks, Harvey, 1972c. Notes on police assessment of moral character. In David N. 
Sudnow (ed.), Studies in Social Interaction (New York: The Free Press), 280-93. 

Sacks, Harvey, 1973. On some puns, with some intimations. In Roger W. Shuy (ed.), 
Report of the 23rd Annual Roundtable Meeting on Linguistics and Language Studies 
(Washington, DC: Georgetown University Press), 135-44. 

Sacks, Harvey, 1974. An analysis of the course of a joke’s telling in conversation. In 
Richard Bauman and Joel Sherzer (eds). Explorations in the Ethnography of Speaking 
(Cambridge: Cambridge University Press), 337-53. 

Sacks, Harvey, 1975. Everyone has to lie. In Ben Blount and Mary Sanches (eds). 
Sociocultural Dimensions of Language Use (New York: Academic Press), 57-80. 

Sacks, Harvey, 1978. Some technical considerations of a dirty joke. In Jim Schenkein 
(ed.). Studies in the Organization of Conversational Interaction (New York: Academic 
Press), 249-70. 

Sacks, Harvey, 1985. The inference-making machine: notes on observability. In van 
Dijk, 1985, 13-22. 

Sacks, Harvey, 1986. Some considerations of a story told in ordinary conversation. 
Poetics , 15, 127-38. 

Sacks, Harvey, 1987 [1973]. On the preferences for agreement and contiguity in 
sequences in conversation. In Button and Lee, 1987, 54-69. 

lxii Introduction 

Sacks, Harvey, 1987. “You want to find out if anybody really does care” In Button and 
Lee, 1987, 219-25. 

Sacks, Harvey, 1988/89. On members’ measurement systems. Research on Language 
and Social Interaction, 22, 45-60. 

Sacks, Harvey, ScheglofF, Emanuel A., and Jefferson, Gail, 1974. A simplest systematics 
for the organization of turn-taking for conversation. Language 50, 696-735. 

Schegloff, Emanuel A., 1963. Toward a reading of psychiatric theory. Berkeley Journal of 
Sociology, 8, 61-91. 

Schegloff, Emanuel A., 1967. The First Five Seconds: the order of conversational openings. 
PhD Dissertation, Department of Sociology, University of California, Berkeley. 

Schegloff, Emanuel A., 1984 [1976]. On Some questions and ambiguities in conversa- 
tion. In Atkinson and Heritage, 1984, 28-52, [first appeared in Pragmatics Microfiche , 
1976, 2.2, D8-G1.] 

Schegloff Emanuel A., 1988. Goffman and the analysis of conversation. In Paul Drew 
and Tony Wooton (eds), Erving Goffman: perspectives on the interaction order (Cam- 
bridge: Polity Press). 

Schelling, Thomas, I960. The Strategy of Conflict (Cambridge, Mass.: Harvard Univer- 
sity Press). 

Schenkein, Jim (ed.), 1978. Studies in the Organization of Conversational Interaction (New 
York: Academic Press). 

Searle, John, 1965. What is a speech act? In Max Black (ed.), Philosophy in America 
(Ithaca: Cornell University Press), 221-39. 

Searle, John, 1969. Speech Acts (Cambridge: Cambridge University Press). 

Searle, John, 1976. The classification of illocutionary acts. Language in Society, 5, 1-24. 

Sudnow, David N., 1967. Passing On (Englewood Cliffs, NJ.: Prentice Hall). 

Sudnow, David N., 1972a. Temporal parameters of interpersonal observation. In 
Sudnow, 1972b, 259-79. 

Sudnow, David N. (ed.), 1972b. Studies in Social Interaction (New York: Free Press). 

Thibaut, John and Kelley, Harold, 1959. The Social Psychology of Groups (New York: 
John Wiley). 

van Dijk, Teun (ed.), 1985. Handbook of Discourse Analysis, volume 3: Discourse and 
Dialogue (London: Academic Press). 

von Neumann, John and Morgenstern, Oskar, 1944. Theory of Games and Economic 
Behavior (Princeton: Princeton University Press). 

Weber, Max, 1952. Ancient Judaism, translated and edited by Hans H. Gerth and Don 
Martindale (Glencoe, III.: The Free Press). 

Zimmerman, D. and West, C. (eds), 1980. Sociological Inquiry, 50, 3-4 [special issue on 
language and social interaction]. 


Beginning with the Fall Quarter of 1968, Harvey Sacks’ lectures were 
delivered at the then recently established Irvine campus of the University of 
California. The spirit of the new campus - at least of its School of Social 
Science under the leadership of its Dean, James March - was quite in keeping 
with the character of the 1960s. It was infused with a sense of possibility in 
its academic and scientific ambitions and was correspondingly innovative in 
organizational form. It dispensed with traditional academic disciplinary 
boundaries and encouraged small groups of faculty to develop new research 
enterprises and to define the terms - and requirements - of graduate degrees. 
The central theme was the unleashing of high quality minds to follow their 
scholarly and scientific instincts wherever the subject-matter, the theoretical 
thrust, or the methodological possibilities seemed to lead, free of the 
constraints imposed by traditional conceptions of disciplinary boundaries and 
other “professional” obstacles to developments which could genuinely 

Whatever elements of his situation at UCLA suggested the possibility of 
leaving, the animating ethos of Irvine’s School of Social Science was very well 
suited indeed to Sacks’ own intellectual metier and character, and to the 
disciplinary iconoclasm of his intellectual enterprise. It was a felicitous 
matching of person and institution. Although Sacks developed a number of 
close ties to faculty colleagues and played a distinctive role in the inescapable 
politics of the academy - politics whose importance was amplified by the 
minimized institutional apparatus of the School - in his work Sacks pursued 
his own course and did not establish sustained collaborative undertakings 
with others on the faculty. This too was a viable possibility within the School’s 
culture. Sacks spent the remainder of his academic career at Irvine, although 
at the very end he was considering another move. 

There is little question that the character of Sacks’ work as it is displayed 
in these lectures (as well as in those of Volume 1) was in various ways shaped 

The introduction to Volume 1 presented some biographical information on Harvey Sacks’ 
education, and set the early phases of his work as presented in his lectures from 1964 to 1968 
in the context of the academic social science of the time. That material is not repeated here, 
and the reader interested in this background is referred to the prior volume. The present 
introduction is concerned less with tracing linkages and contrasts between Sacks’ work and 
other developments in social science (although there is some discussion of this sort) and more 
with the treatment of Sacks’ work in its own terms. 

I am indebted to Paul Drew and to John Heritage for reading a draft of this introduction 
on my behalf, and for the collegiality and helpfulness of their responses. 



both by the larger social and cultural Zeitgeist of America in the 1960s 1 and 
by the specific local ambience of southern California during that period, 
within which the scene at Irvine played itself out. A delineation of those 
connections will have to await another occasion. But there is equally little 
question that Sacks’ oeuvre cannot be reduced to the socio-cultural environ- 
ment in which it happened to emerge. The distinctiveness of his vision was 
formed before the 1960s, and his pursuit of a distinctive path antedated that 
special time as well. And it was formed not only in California but in such 
bastions of academic tradition as Columbia, Harvard and Yale. 2 

There is much continuity between the lectures published in Volumes 1 and 
2. Most notably, the extraordinary, detailed analyses of small bits of 
conversation in which whole social worlds and whole ranges of personal 
experience are dissected from out of apparent interactional detritus continue 
to be interlaced with more abstract theoretical and methodological discus- 
sions. Various substantive themes persist as well - most importantly the 
preoccupation with sequential analysis, and a continuing tacit preoccupation 
with how to conceive of “culture.” 

There are discontinuities as well. Topically, discussions of membership 
categorization devices per se are not resumed, although on occasion the 
resources of that body of work and the problems attendant on “doing 
formulating” figure centrally, for example, in the lectures for Winter 1969. 
A concern with storytelling in conversation which first emerges in the Spring 
1968 term is much more fully developed, beginning with considerations of 
sequential organization but extending into quite new analytic directions. 
Observations about sound patterning and other “literary” aspects of word 
selection emerge for the first time, and are taken up in several of the lecture 

Thematic and analytic continuities and innovations aside, there are some 
new stances taken up in the lectures published in Volume 2 to which it may 
be useful to call attention, if only briefly. Some of these may serve to suggest 
connections between the several sets of lectures which compose this volume; 
others may serve as ways of focussing an initial orientation to each set of 
lectures in turn. I begin with a theme which first appears in Fall 1968 but 
recurs thereafter. 


One apparent shift of stance which appears concomitant with the move to 

'Recall, for example, (as a Los Angeles commentator recently did) that among the events 
of just 1968 were counted “the year of McCarthy for President, the fall of L.B.J., the 
assassinations of King and Robert Kennedy, the Beatles’ White Album, Motown and “2001: 
A Space Odyssey”; of war, orgy and dreams of peace, in the summer after the Summer of 
Love” (Los Angeles Times , August 23, 1991). 

2 Views which reduce work like Sacks’ to something like the product of a California flower 
child (for example, Gellner, 1975) are not only demeaning and intellectually evasive in 
dismissing by epithet what they cannot decisively engage in substance; they are factually 
ill-informed as well. 



Irvine is a turn toward systematicity and toward the relevance of substantial 
amounts of data, that is, aggregates of conversations or of instances of 
particular phenomena in it. Consider, for example, the stance which Sacks 
adopts in launching his discussion of turn-taking in lecture 3 (and continuing 
into lecture 4) for Fall 1968. Among the key points in this new stance are the 
following (all from lecture 3, p. 32): 

What I want to do is to lay out in as general a way as possible at this 
point how the sequential organization of conversation is constituted. 

Note the shift to “general” and the generic reference to “conversation.” 

I start out with two observations about single conversations. . . 

Note that although the observations are about “single conversations,” they 
are about aggregates of them. 

I give in this first instance no materials for the observations, in that they 
are grossly apparent. 

Note this shift in practice; nothing in particular is the point of departure; an 
observation about a regularity in an aggregate is the point of departure. 

By the term ‘grossly’ I mean that while they’re overwhelmingly present 
features, they are also sometimes not present features - and their 
sometimes non-presence is something I will talk to at considerable 

Note that the issue here is the dealing with occurrences that depart from a 
general practice, “sometimes non-presence.” 

The shift, then, is to: 

an order of organization , rather than a particular practice, of talking; 

a class of places in an aggregate of data , rather than an excerpt; 

an organizationally characterized problem or form of interactional work, rather 

than an individually designed outcome; 

invariancies of features rather than context-specified practices. 

This is not, of course, a total shift of procedure. In lectures 5 and 6 Sacks 
again presents particular materials, and explores turn-taking issues (among 
others) in the context of a developing set of observations about that excerpt. 
But there are readily observable consequences of this shift in point of 
departure and analytic stance; one of these is an increasing (or increasingly 
explicit) orientation to organization and structure in the domain of conver- 
sational conduct. Again, discussion must be limited. 

Near the beginning of lecture 4 (p. 44) Sacks develops the point that there 



are grounds, built into the organization of conversation, for listening to every 
utterance for any participant willing to speak if selected, or willing to wait if 
another is selected to speak. The upshot: 

So again there’s motivation to listen, which is independent of any rule 
that would say ‘you ought to listen in conversation;’ motivation to listen 
which turns on a willingness to speak or an interest in speaking. 

This is not first time this point has been made in these lectures, but it has a 
different resonance in the context of the new casting of turn-taking. What is 
its interest? 

Let us note first that the core point seems to be a grounding of listening to 
utterances in the technical requirements of talking or not talking, the technical 
requirements of the organization of proper conduct in that regard, rather than 
its grounding in a normative injunction directed to that outcome specifically. 
That is, an analytic concern for parsimony is at work here, setting aside a 
normative constraint, a “rule” if you will, designed specifically to secure 
“listening” or “attentiveness,” or showing that such a rule, if there is one, is 
not there solely to secure attentiveness, because attentiveness is already a 
natural, a technical, by-product of the organization of turn-taking . 3 

Now this is surely not to deny a normative component to the organiza- 
tion of interaction or conversation; surely, the sorts of mechanisms by which 
the turn-taking organization is constituted are normative in character, for 
the participants and consequently for analysts . 4 On the other hand, it does 
seem to subordinate considerations which might be termed “politeness” to 
ones which might be termed “technically constitutive” or “sequence 
organizational.” The parsimony considerations here seem to take the form: 
what sort of basic organization would both drive the prima facie organiza- 
tion of the talk and engender whatever auxiliary effects seem to be 

One implication is that listening is not vulnerable to (or is less vulnerable 
to) whatever it is that may weaken persons’ commitment to observe 

’Subsequently, in lecture 2 for Spring 1972, pp. 535-7, Sacks returns to this theme in the 
context of his discussion of adjacency pairs, the virtually unrestricted freedom of occurrence of 
their “first pair parts,” and the potential usability of first pair parts for selecting a next speaker. 
From this, Sacks observes, it “falls out” that a participant willing to speak if selected to do 
so will have to listen to everything said, for at any point a first pair part selecting them may 
be done. This account in Spring 1972 is different only in its focus on adjacency pairs as 
instruments of next speaker selection, rather than on the tum-taking organization per se. 

4 And, indeed, at Fall 1968, lecture 4, p. 50, Sacks proposes, "We have in the first instance, 
some formal normative features for conversation, which are in a way a public law for 
conversation: One party at a time. . .” etc. 

’Compare here the discussion (introduction to Volume 1, p. xxviii-xxix; 1 — li) of the 
treatment of “indirect speech acts” with primary respect to considerations of politeness and 
sequential organization respectively. 



normative constraints. 6 The point here is that the basis for listening is not as 
much at risk as injunctions to be polite, when violations of politeness had 
become, for example, a systematic political tactic on university campuses. 
Listening was grounded in self-interest (wanting to talk, or being willing to) 
and the technical requirements of implementing it. Departures from “features 
of conversation” should be understood, therefore, not so much by reference to 
motivated deviation from rules prescribing them as by reference to modified 
operation of the system of which they are a by-product - for example, in 
response to variations in context or transient problems in internal coordina- 

Note the bearing of this tack on the claims of certain forms of “intention- 
alist” theorizing (such as those of Searle, 1991) that our knowledge of human 
action or conduct has only been advanced when “patterns” (as Searle calls 
them) can be shown to be the causal products of intentions to produce them. 
If the stance taken here by Sacks is correct, then observed distributions of 
attention (i.e. , observed patterns of listening to others’ ongoing talk) may best 
be understood not as the product of an intention to comply with a rule 
mandating such attention (even if there was such a rule), but as an imposed 
requirement for achieving such outcomes as talking if asked to, or withhold- 
ing talk if another is asked to. (For a more general statement of this theme, 
see lecture 2 for Spring 1970, and the discussion of that lecture below at 
p. xxiv and n. 17). 

Sacks’ grounding of the organization of attention/listening in the individ- 
ual participant’s willingness to talk if asked to or to remain silent (even with 
something to say) if another has been selected to talk, itself embodies a 
distinctly sociological theme in accounts of social order. Developed in Sacks’ 
account of turn-taking most explicitly at lecture 4, pp. 50-2, this theme 
understands the enforcement of the turn-taking organization to work by its 
identification with individual participants’ rights and interests. So understood, 
individuals are mobilized to defend their rights and interests (e.g., their turn 
space); the emotions are recruited to this enterprise as well, such that 
violations of “one-at-a-time” become treated as invasions of some speaker’s 
right, and that incursion engenders anger in defense of those rights, that 
emotional energy being put in the service of a socially organized enforcement 
mechanism for the turn-taking organization. Further, gossip, reputation, and 
the like can be recruited into that enforcement mechanism as well, e.g., under 
the aegis of violators being “rude.” This, then, is how this class of violations 
gets seen as violations of “politeness,” and it is in this light that we should 
understand at least some “politeness” considerations. That is, it is by reference 
to “politeness” that sanctioning is vernacularly formulated, while the actual 
occasioning of the violations may be less a matter of normative etiquette 
and its violations, and more a matter of technical organization or action 
implementation, effectuated through the identification of individuals’ 

Something which was, of course, increasingly remarked upon in the 1960s, and certainly 
not less in southern California than elsewhere. 



rights/interests with the resource which the turn-taking organization 
distributes. 7 

Throughout this discussion, it is apparent that considerations of systema- 
ticity, structure and organization play an important role in understanding 
orderly conduct observed across aggregates of data. 8 Although not all of the 
Fall 1968 lectures display this stance, it does play a continuing (even an 
increasing) role in Sacks’ subsequent work, including subsequent lectures, for 
example, the lectures of Spring 1972 on adjacency pair organization. 


Although “turn-taking organization” is the substantive focus for the Fall 
1968 lectures, Sacks does not begin the course with a lecture on that topic; 
indeed, he does not begin his discussion of it until lecture 3. The first two 
lectures present another “take” on the “second stories” theme first treated in 
the previous spring, at UCLA, and it may useful to linger for a moment on 
what Sacks was doing in starting this course the way he did. 

Note that the first lecture announces that it will be concerned with 
something other than what Sacks otherwise plans to focus on. He begins: 

Hereafter I’ll begin with some rather initial considerations about 
sequencing in conversation. But this time I’m going to put us right into 
the middle of things and pick a fragment that will introduce the range 
of things I figure I can do. 

He does this, he says, in order not to stake his claim on the usual insignia of 
academic work (“ . . .its theoretical underpinnings, its hopes for the future, its 
methodological elegance, its theoretical scope . . .”), but on the “interesting- 
ness” of the findings. This was a task which Sacks set himself in the late 
1960s - to have “bits” with which to tell lay people (including, for this 

7 The theme of ensuring outcomes by identifying them with individuals’ property, interests 
or rights - a familiar theme in certain '‘liberal” traditions of social theory - comes up again 
in a strikingly different context in Sacks’ treatment of the motivated preservation of 
experiences in memory for later retrieval and telling (cf. Spring 1970, lecture 5, pp. 257-9, 
and below at pp. xxv-xxvi). 

8 Another kind of consequence of this new stance, especially with respect to asserting claims 
about aggregates of data rather than specific data fragments, is an occasional vulnerability in 
the grounding of some claims in these lectures. Without materials as a shared point of 
departure, it is at times unclear what actual things Sacks is talking about, and, therefore, how 
to assess what he is saying. There are assertions, when the work takes this form, about things 
which are said to happen "all the time,” which may not seem all that familiar to the reader. 
(E.g., for this reader, p. 49: "Some people say about each other, ‘Why is it that we can never 
have a conversation without it ending up in an argument?’ And in that it is a thing that is said 
all the time, it is of interest to see how it could be sensible.”) Of course, what Sacks asserts 
- at times ex cathedra - and the tack which he takes, regularly turn out to be of great interest 
for their strategy of analysis even when subject to such reservations. 



purpose, other “straight” academics) what “the work” consisted in which 
would have a kind of transparent appeal and interest, readily presentable and 
graspable in a relatively non-technical way, capturing “experiences” virtually 
anyone would have had access to more or less directly, etc. 

For a while, a regularly offered “for instance” was what Sacks proposed to 
be an exemption from the ordinary recipient-design “rule” or “practice,” for 
(among other forms of talk) storytelling - “Don’t tell others what you figure 
they already know.” Sacks proposed that there is an exemption for spouses. 
This is to be understood as a practice coordinate with a mandate to tell 
spouses many things first, before they are told to anyone else. Then, given that 
spouses are present together on many interactional occasions and that each 
would have been first to be told most tellables, without the exemption many 
tellables would have major constraints on their subsequent reliability to 

But the exemption engenders its own troubles. Because spouses’ presence 
need not deter re-tellings, spouses may find themselves having to hear the 
same stories over and over again. And the presence of an already “knowing” 
person can have consequences for the form that the telling takes. As a result, 
there is a pressure for the separation of spouses in social occasions where these 
various cultural practices and orientations are in effect (thus, for example, 
rendering them free for groupings based on other features, e.g., gender). 9 

This was a neat little package, in which a familiar social experience did 
seem readily traceable to practices of talking which ostensibly had little to 
with them (or with anything of general interest), and was appealing and 
satisfying as an “illustration” of the work. 

Much in these first two lectures has the flavor that would make it attractive 
on these grounds. Especially points well into the discussion of lecture 1, 
regarding the counter-intuitive relative paucity of “things to talk about” with 
those one has not talked to in a long time as compared with the ready supply 
with those one talks to daily, 10 are just the sort of thing that Sacks saw as 
useful in these ways. His departure from his planned theme in the initial 
lectures in order to do this repeat “take” on second stories may embody his 
treatment of the class members as part of a larger general public whch had to 
be appealed to, at least initially, on the grounds of common experience. 1 1 

One other aspect of these lectures which occurs in various of the sets but is 
striking in the Fall 1968 set is what I will refer to as an aspect of their rhetoric. 
One form which this rhetoric takes is the assertion, after some particular 

9 A version of this line of analysis appears in this volume as lecture 4 for Fall 1971, where, 
however, it is touched off by a particular data fragment, from which Sacks formulates the 
problem of spouses’ talk. 

ll> This theme is returned to in the initial lecture for Winter 1970, p. 172. 

n In the lectures for Spring 1970, Sacks is explicit about the special cast being given the 
first lecture. Strikingly, the topic which here in the Fall 1968 lectures serves as the accessible 
beginning for the course becomes in the Spring 1970 lectures the “much more severely 
technical” (Spring 1970, lecture 1) material which warrants a more accessible introductory 



analysis or type of analysis has been offered, of its “normality” as a scientific or 
disciplinary development. So, in Fall 1968, lecture 3, p. 38, Sacks proposes: 

In its fashion the history I’ve recounted is a perfectly natural history; i.e., 
it would be perfectly natural for whatever course of development of 
analysis of something that what you’re looking for initially when you 
look at something - a plant, a social object, whatever it may be - is to 
find some parts. One would begin off, then, with things like ‘greetings’ 
and in due course come to things like ‘one at a time’ and ‘speaker 
change’ occurring. 

Now, Sacks had read considerably in the history and philosophy of science, but 
the claim made here is merely asserted and not developed by reference to that 
literature. And what is asserted is an actual course of events of Sacks’ own 
making, transformed into a putative generalized course of events which con- 
stitute normality or “natural history.” The inter-convertability of modalities 
such as instructions and historicized descriptions is something Sacks was well 
aware of. It is a way of subsuming new departures, and a position staked out 
without benefit of colleagues close by, under an umbrella of “normal science.” 
Again at Fall 1968, lecture 4, pp. 54-5, Sacks invokes “naturalness”. 
Having made a point about the co-occurrence of ‘one at a time’ and ‘speaker 
change recurs’ as features of conversation that are “basic,” he then gives an 
argument for this basic-ness (i.e., that the system is self-organizing, in that 
breakdowns/violations are organized by reference not to some other rules but 
by reference to these very same ones). 12 And then: 

And I take it that that’s an extremely natural criterion for some rules 
being basic; that is to say, when you reach them, you reach the ground. 
There are no other rules which deal with how to deal with violations of 

It seems clear that this is not offered as an account of some actual history of 
usages of “basic,” but as an effort to put into perspective the status of what 
he was proposing. Here the rhetoric of “naturalness” is “aggressive,” in 
claiming a status within some putative developmental course of a discipline. 
Elsewhere, a more “defensive” (though hardly apologetic) tack is taken, as, 

'“A similar argument is made with respect to adjacency pair organization in Spring 1972, 
lecture 2; cf. below pp. xliv-xlv. 

The contrast, it may be useful to mention (or one contrast at least) to this “self-organizing” 
property is the sort of feature taken up in the 'Two preferences . . .’ paper (Sacks and 
Schegloff, 1979), which is concerned with “second order organization.” There, if two features 
meant to co-occur (in that context, “minimization” and “recipient design;” as here, “one at 
a time” and “speaker change recurrence”) are not combinable on some occasion, there is an 
extrinsic procedure for reconciling the conflict, i.e., relaxing one feature until the other can be 
achieved. The parallel argument for “interruption” (as an instance of non-combinability of 
“one at a time” and “speaker change”) being resolved in a “self-organizing” fashion has yet 
to be presented formally. 



for example, in lecture 4 for Spring 1972, where Sacks offers an aside while 
launching a discussion of adjacency pairs by formulating three abstract 
utterance positions in conversation - “last, current, and next utterance.” He 
says (pp. 554-5), 

A lot of this will sound awfully banal but it’s far from that, so you’ll 
have to jolt yourself - if I don’t jolt you - into thinking that it’s not, 
after all, something anyone could have said; it’s not that it’s nothing; it’s 
not that it has no consequences. 

This should be appreciated as being at least as much self-directed as addressed 
to the audience - either the physically co-present class or the audience 
wherever. It is a sort of girding of loins before battle; a sort of assertion of 

In the intermissions and aftermaths of days we were working together, 
Sacks used to bemoan the difficulty of the work. One of his metaphors for it 
was the need to be able to “look around the corner of the future,” that is, to 
be able to see ahead to that formulation of the organization of the world 
which would appear in retrospect to have been obvious. And often this 
seemed to turn on seeing in some (but not other) apparently commonsense 
characterizations of empirical objects their potential for carrying heavy and 
complex theoretic/analy tic loads. One problem which this posed was the 
vulnerability to lapsing back into a mundane, vernacular, commonsense 
hearing/understanding of those terms - one which would not sustain the 
analytic load they were to carry, but would reduce to some “banal” 
pre-theoretic assertion. It is that sort of vulnerability - both in his audience 
and in himself, however differently for each - that this invocation seems 
designed to confront; and it is similar vulnerability and transient self-doubts 
which the “natural development” rhetoric seems designed to combat. 


The Winter 1969 lectures presented here do not themselves compose a 
thematically organized set, or even several such. Rather, they present a variety 
of analytic topics and problems occasioned by efforts to come to terms with a 
single stretch of material taken from the first of a series of group therapy 
sessions with “adolescents” which Sacks had recorded (and, later in the 
course, other materials as well). Although some considerations raised in 
dealing with one part of this excerpt may come up in connection with another, 
these lectures do not appear to have been designed to constitute coherent, 
systematic treatments; still, in some instances (e.g., lecture 3) they do seem to 
come together quite nicely. For the most part, however, some fragment of the 
data segment is isolated for treatment, and then several sorts of interest in it 
are extracted and addressed. 

Not that this detracts from the striking and unexpected lines of analysis 

xviii Introduction 

which Sacks develops from his materials in the various, largely independent 
discussions. The tone is set from the very beginning. 

In the Spring 1966 lectures Sacks had examined the notion of “posses- 
sion,” and in various respects reconstituted what sort of a cultural artifact it 

is. In lecture 1 for Winter 1969 (although not explicitly related to the earlier 
discussion), he makes another sort of novel use of “possession” or “owner- 
ship.” In discussing the noticing/remarking by one participant in the group 
therapy session on the hole in another’s shoe, Sacks notes that that the shoe 
is owned by its wearer may entail that another cannot take it, but it does not 
entail that another cannot talk about it. Further, if another talks about it, it 
is very likely that its owner will talk next, or soon. So “ownership” is 
conversationally consequential. 

Furthermore, one of the generic matters conversation is centrally taken up 
with is the things that the participants have brought with them to the 
conversational occasion - their clothing, possessions, bodies, events they enact, 
etc. The talk works off what the parties have brought; and parties can then 
bring what they bring in part by virtue of the talk that may be made about 

it. And persons may avoid being present to a conversation by virtue of what 
they must necessarily bring to it (e.g., the current state of their bodies, 
possessions, etc.), in view of the talk which that company is likely to make 
about it. Possessions are then relevant not only to “the economy;” they are 
central to the “conversational economy” as well. And “ownership” turns out 
to be a social/sociological category which is consequential in hitherto 
unappreciated respects. 

There are other sociological threads running through many of the 
discussions in these Winter 1969 lectures. One such theme concerns group 
formation, membership claims, and different ways of “partitioning a 
population” 13 to find who belongs together and who not. As the last of these 
clauses may suggest, it is by way of interactants’ deployments of membership 
categories and ways of identifying or formulating one another that these 
various topics are addressed. In lecture 2, the issue is posed by how someone 
is praised without impugning the status of the others (the issue being who is 
the same category with the praised one and who not). In lecture 3 it is the 
alternative ways of grouping two of the attendees of the therapy sessions - 
Roger and A1 - together vis-a-vis the observer, as between patient/observer 
and performer/audience. In lecture 7 it is the issue of who is a “hotrodder” 

13 By “partitioning a population,” it may be recalled, Sacks refers to the results of 
formulating a set of persons by reference to the categories in some empirically coherent set of 
categories, i.e., categories which compose “a set” in an empirical sense. “Partitioning 
constancy” (lecture 3, p. 110) describes the outcome when a same collection of persons are 
distributed in the same way by reference to two or more different sets of categories. Thus, later 
in this paragraph of the text, the category sets “patient/observer” and “performer/audience” 
divide up the co-present persons in cognate fashion - the ones who are co-members of the 
category “patient” in one set of categories being co-members of the category “performer” in 
the other; these category-sets then display partitioning constancy for this population of 
persons, or constitute “analog structures,” as Sacks also refers to the matter there (ibid). 



(or “hippie”) and who not, who is “authorized” to make such a judgement 
and how some persons “patrol the borders” of the category. In lecture 8, the 
issue is posed by reference to alternative ways of seeing some collection of 
persons in some place as legitimate or not, via their alternative formulations 
as “gals and guys” or “den mother.” 

Another, more methodological, theme which informs a number of the 
lectures across considerable variation in substantive topic concerns the 
relationship between “intuition” and “formal analysis” on the researcher’s 
side on the one hand, and the relationship between analytical “formality”/ 
“abstractness” in contrast to the “concreteness” of “lived experience” for the 
“ordinary actor” on the other. 

Sacks’ characterization of what he is doing in lecture 2 - on “safe 
compliments” - is instructive; its logic here echoes that of the analysis of 
“invitations” as early as Spring 1966. In a discussion initially targetted at 
“the weather” as a “safe” topic, Sacks begins elsewhere: 

I did some work on ‘compliments,’ specifically on what I called ‘safe 
compliments,’ the idea being to see what it was about some compli- 
ments that made them ‘safe’ compliments, i.e., to turn an initial 
observation into an analysis . . . The question then is, can we extract 
from the sort of thing [some particular compliment} is, a set of features 
which will locate a class of compliments like it, which are also safe 
compliments? Where that is a test of the fact that we had some 
generative features, [emphasis supplied} 

Then, after developing an analysis of what makes one class of compliments 

Now the question is, with respect to ‘weather talk’, what do we need, 
to be able to show that ‘the weather’ is a ‘safe topic’? What we need is 
to develop a notion of ‘safe’ for topics so that we can have said something 
when we say ‘weather’ is a safe topic.' The discussion on ‘safe 
compliments’ was to give a sense that something could be done with a 
notion of ‘safe’, something of a formal sort , i.e., it doesn’t have to be merely 
an intuition , but what’s involved in something being ‘safe’ can be laid 
out. [emphasis supplied}. 

Now it should be clear from this treatment that what the professional analyst 
might come to analyze as the formal features that make for “safeness” - 
whether for compliments or for topics - is proposed to be “real” for parties 
to talk-in-interaction; it is for them, after all, that it is proposed that the 
“safety” matters, and they who may suffer from the lack of it. Still, such 
formal accounts are vulnerable to charges of “formalism,” of imposing 
analysts’ categories onto the lived experience of the participants, and the like. 
To this theme it is useful to juxtapose Sacks’ discussion in lecture 3 of one way 
in which two of the “patients” in the group therapy session deal with the fact 



(of which they have been apprised) that there is an observer behind a one-way 
mirror in the room. They “enact” a scene of “personnel just before a 
performance,” calling out “Testing, one two three” and the like. It is in this 
regard that Sacks points out the “partitioning constancy” in that setting 
between “patient/observer” and “performer/audience,” which allows the 
latter set of categories to provide a set of “cover” identities, at least 
transiently. 14 

Now this appears to ascribe to the teenage therapy patients a kind of 
abstract or formal analysis of their circumstances which may appear to violate 
our understanding of their lived experience. But Sacks argues (lecture 3, 
pp. 1 10-11) that what is at issue in using a “theater” frame to deal with the 
presence of an observer is that people 

have their circumstances available to them in an abstract way, such that 
they can use the abstract characteristics of their circumstances to locate 
other circumstances that stand in a strong abstract relationship to their 
current circumstances. 

The relevance of this point is precisely to counter the objection to this whole 
direction of analysis that, in explicating underlying abstract or formal features 
of ordinary activities, violence is done to the lived-experience of those activities 
for the actors who engage in them. By contrast, Sacks is proposing here that 
part of ordinary Members’ competence is specifically an abstract understand- 
ing of their circumstances and activities, an abstract knowledge drawn upon 
in constructing further courses of action, and usable to construct further 
courses of action in a fashion coordinated with others. Thus: 

How can they use that abstract knowledge? They are able to use such 
knowledge to locate circumstances which have features that stand in a 
strong relationship to the initial circumstance, and those features are 
then used to project actions by reference to those other circumstances, 
which actions have some hope of being picked up. It’s not just one 
person who is by himself capable of that, but he can have hopes that 
others can see what he’s doing, see it fast, and collaborate with him. 

The transformation by analysts of intuition into “something of a formal sort” 
is thus not merely a requirement of disciplined inquiry, its results are 
themselves meant to capture features of the procedures by which ordinary 
conduct by ordinary members is methodically achieved. 

14 See the earlier treatment of this episode in Volume 1, lecture 14 for Spring 1967. 

Aside from the focus which the text brings to this discussion, Sacks’ demonstration of what 
might be involved in seriously grappling with the effects which observers might have on a 
“scene being observed’’ is a salutary one in refusing to settle for a simple and cliched concern 
about “Heisenbergian” influences of observation itself. Rather, it insists on a detailed attention 
to how and what sorts of changes in conduct there might be, how they are to be understood, 
and how they would/might bear on what an observer makes of that conduct. 




The lectures for Winter 1970 begin with a focus on the overall structural 
organization of the unit “a single conversation,” linger in lecture 2 on the 
theme of exploiting whatever topics come to notice in the intensive exami- 
nation of a single conversation, and then return to considerations of overall 
structural organization. There is much here that is penetrating and revelatory, 
concerning such objects as ‘‘the reason for the call” and “reason for the call 
relationships,” as well as “no reason for the call calls,” and relationships built 
on them, to mention only some of the attractions of the first lecture. 

There are elements in these lectures whose relevance is related to some of 
the new emphases which I earlier suggested inform the lectures starting in late 
1968. I want to take note in particular of a passage of two to three paragraphs 
at pp. 168-9 of lecture 1 in which a theme first appearing in Fall 1968 
reappears, and that is the relevance of examining a fragment from a 
conversation in the context of (or juxtaposed with) other products of the sort 
of “machinery” conjectured to be involved, other instances of the “same sort 
of thing;” that is, the use of aggregates of data. In Fall 1968 this theme 
surfaced in passing with respect to turn-taking; here it comes up in a more 
sustained way with respect to the openings of conversations, both (and 
especially here) on the telephone and in co-present interaction. 

This is a topic - single case analysis versus working with collections of data 
- which is not uncontroversial, and which Sacks and I discussed at 
considerable length over the years. This is not the place for a thorough airing 
of the issues or of those discussions. The key point here in Sacks’ treatment in 
lectures 1 and 2, however, is that a proper grasp of what might be going on 
in a conversational opening in some particular setting might require a grasp 
of the range which the “machinery” involved in the production of the 
phenomena involved could produce, and this might require examination of a 
considerable array of data. 15 

Once dealing with an array of data taken to be “comparable,” a 
comparative analysis may appear to be needed, and this can itself give rise to 
some methods of analysis which may obscure how the material being studied 
may have been produced, rather than illuminating it. One such analytic 
procedure requiring considerable care and reflection is “format-and-slot 
analysis,” in which the prototypic problem is cast as a selection among 
alternative terms which could be used for a same reference, or alternative 
items which could be employed at a certain juncture in the talk, a juncture 
formulated by the format of the talk in which it is embedded. It is not that 
this form of analysis is flawed in principle; conversation analytic treatments of 
reference - reference to persons, to places, etc. - have exploited it. 

Sacks points out, however, that there are circumstances in which alterna- 
tives to a term actually employed would/could not be used, even if they were 
“correct. ” He takes as his case in point a telephone call in which the caller has 

15 See the discussion below at pp. xxxix-xl and n. 28 . 

xxii Introduction 

called her friend about a commotion which was observed at the friend’s place 
of work, a department store called “Bullock’s.” Sacks argues that in 
proposing that one has called to tell another “what happened at Bullock’s 
today,” the reference to “today” is not incidental. It is not properly 
understood as being selected from a set of cognate temporal references. Had 
the event happened several days earlier, the caller would not offer that 
temporal referent in the same utterance format; she might not tell the story, 
or find it tellable, at all. Indeed, given that the caller called in order to tell the 
story, she might not have called at all. For it is its occurence “today,” Sacks 
proposes, which makes the event “news,” and thus a possible “reason for the 
call,” and hence in first topic position in the call. It is the fact that it was 
“today” that makes for a temporal reference being used at all, rather than a 
temporal reference being somehow slated to occur, with a selection procedure 
then invoked to find the term to be plugged into that slot. And further, it is 
not that its occurrence “today” makes it tellable as news per se; it makes it 
tellable to one with whom the teller talks daily. It might not be tellable to a 
twice yearly interlocutor, even if it happened “today,” for it may not have the 
stature to be told in a six-monthly conversation. So all of the discussion is itself 
subject to considerations of recipient design. These widening ripples of 
analytic consideration surround the use of “slot-and-format” analysis, and 
may render its invocation questionable. 

Lecture 2 for Winter 1970 (at pp. 184-7) contains what is to my mind 
one of the most striking discussions in all the lectures. Here Sacks turns a 
seemingly technical dissection of the mundane story mentioned above - about 
the commotion outside a department store told by one friend to another - into 
a stunning demonstration of the alternative grasps of a scene which may 
present themselves to different sorts of viewers - Sacks refers to it as having 
become “kind of a distributional phenomenon.” 

His account begins with the contrast between the actual teller’s perception 
that there- was- trouble-and-the-police-were-taking-care-of-it on the one hand, 
and, on the other, what Sacks proposes others (e.g., residents of the “ghetto”) 
might see as there-being-trouble-and-the-police-were-engendering-it. He pro- 
ceeds through a series of further related observations, for example, the 
assuredness of the actual observer that her position as uninvolved witness is 
unquestioned, as compared with the possibilities which other categories of 
person finding themselves on such a scene would be required to entertain and 
protect themselves against - for example, the possibility that they would be 
treated as accomplices in whatever wrongdoing was suspected. The effect is to 
render the scene which the story is intendedly about as equivocal as the 
duck-rabbit of Gestalt psychology, and the actually told story as a situated, 
perspectival version of it. 

Sacks’ observations here carry the conversation-analytic treatment of an 
ordinary story told in conversation to an intersection with traditional themes 
of social and political analysis, and can well have served as a revelatory 
component of a liberal arts education for white middle-class undergraduates 
in Orange County, California in the aftermath of the Watts riots in Los 

Introduction xxiii 

Angeles just to the north, in years which were, in all but their numerical 
depiction, still part of the 1960s. 

This intersection with, and transformation of, vernacular understanding is, 
I would like to stress, not a time-out from technical analysis but a product of 
it. Sacks’ discussion here should be juxtaposed with his discussion of “viewer’s 
maxims’’ in the paper ‘On the analyzability of stories by children’ (1972b) (or 
the first lectures for Spring 1966 on which it was based), where the technical 
basis for these observations may be seen to have been rooted. 


The set of lectures for Spring 1970 is as coherent and stunning in its range and 
perspicacity as anything in the collected lectures. It is the richest single set of 
materials on Sacks’ treatment of storytelling in conversation, and surely 
central to our understanding of stories more generally. 

Here as before (cf. lectures 1 and 2 for Fall 1968) Sacks announces the 
opening lecture 16 as one intended to appeal more broadly to the class than the 
material to follow, which he characterizes as “much more severely technical 
than most people could possibly be interested in.” He continues here the 
practice of developing materials which could give “outsiders” a sense for this 
work and its possible payoffs in a relatively vernacular way. The “more 
accessible” materials of the Fall 1968 lectures, however, had become “much 
more severely technical” by Spring 1970 (at least they were going to be 
presented that way), and now were given their own, more readily accessible, 

Whatever the long term relationship of ethnomethodology and conversa- 
tion analysis turns out to be, this lecture as much as any other in the corpus 
of Sacks’ lectures (at least those to which we have continuing access) 
exemplifies a convergence of the animating impulses of ethnomethodology 
and conversation analysis in its preoccupation with the “ordinary,” the 
“normal,” the “mundane” as achievements. 

With lecture 2 Sacks begins the treatment of stories told in conversation. 
It is a beautifully organized and accessible account of the sequential problem 
of storytelling in conversation by reference to the organization of turn-taking 
in conversation, and the understanding of the “story preface” by reference to 
it (material later presented in Sacks, 1974). Perhaps two points may be 
underscored here which might be overlooked in a reading of the lecture for 
the aforementioned focus. 

The first is Sacks’ self-conscious attention to theorizing as an activity. He 
begins here - as he does in many other lectures - with what he calls an 
“utterly bland fact,” one whose telling surely is not in itself of interest. The 
point, he remarks, is what can be made of such a bland fact. But many bland 

16 This lecture - supplemented by excerpts from lectures 2 and 4, and lecture 1 for Spring 
1971 - has previously appeared in print (Atkinson and Heritage, 1984: 413-29). 



facts lead to nothing beyond themselves. It is necessary, then, to have found 
and pursued such a bland observation as allows something to be made of it. 
In the end, then, the blandness or “obviousness” of some observation is 
neither grounds for ignoring or suppressing it nor, in itself, for asserting it, but 
for seeing if its achievement or consequences can be seen to be more telling 
than the observation itself. 

Such concerns with “theorizing” appear recurrently in this set of lectures 
(as they do in the corpus as a whole). To cite but one additional instance, in 
lecture 3 Sacks remarks on the common practice in everyday life that persons 
take note of “coincidences” - for example, that they rarely go to some place, 
and their interlocutor rarely does, and that it was a coincidence that they both 
happened to do so on the same day and encountered one another there. He 
then proposes: 

I want to see if we can get at the beginning of an answer to how we come 
to see these coincidences. The interest in the beginning of an answer is 
not so much in whether it’s an answer - I don’t have any idea whether 
it’s an answer - but in some way that the answer is built. 

As with the blandness of the point of departure in lecture 2, the concern here 
is with the ways of building an account, of theorizing in the presence of data 
per se, rather than with the final assessment of the adequacy of the account. 
By the end of the lecture, Sacks is again proposing that much of the 
observable orderliness of the world may be better understood as the 
by-products of ambient organizations which are quite unconcerned with these 
outcomes, rather than as products which were the design target of some 
organization. 1 7 

In passing Sacks here produces an account of the perception of coincidences 
that makes of it not a mistaken commonsense notion of probability, but 
something like Marx’ notion of alienation; 18 that is, that persons’ own 
activities (here the practices by which stories are formed up) produce a result 
(an account of activities that is designed to make for relevant-at-that-moment 
tellable stories), which is then perceived not as a product of the design of 
storytelling, but as an independently encountered - and somewhat mysterious 
- “external” reality. 

Additional discussions of this explicitly methodological sort in the Spring 
1970 lectures include an interest in “doing provings” (lecture 5, pp. 25 Iff), 
“getting ... a problem” (lecture 6, pp. 267ff) and the relationship between 
a “sophisticated lay observation” and more technical treatments (lecture 7, 
pp. 271-2). 

17 This is, then, a more general statement of a theme raised in Fall 1968, lecture 4, where 
“listening in conversation” was treated as a technical requirement and result of the operation 
of the turn-taking system, quite apart from any normative regulation explicitly concerned with 
“listening in conversation.” Cf. that lecture, and the discussion above at pp. xii-xiv. 

ls For example, the account of “alienated labour” in the Economic and Philosophical 
Manuscripts (cf. McLellan, 1977: 77-87). 



A second point worth lingering on is Sacks’ treatment of the term “story.” 
Especially in the years following these lectures there has been an explosive 
growth in interest in, and writing about, stories, narrative, “narrative logic,” 
etc., with whole fields and sub-fields (e.g., “narratology”) addressed to this 
subject-matter. Unsurprisingly, the growth of this academic and literary 
industry has spawned a profusion of definitions of the focal object — such as 
“story.” Sacks parries the issues of “what is a story?” and “is this a story?” by 
asking not whether the label “applies” (i.e. , is “correct”), but whether it is 
relevant - that is, relevant to the participants in producing the stretch of talk 
in and through which the object in question was produced. The issue is thus 
transformed from an “external analyst’s” issue into a “a Member’s issue:” 
how does it matter to the teller and the recipients that the talk being produced 
(i in the course of producing it) is “a candidate story”? Lecture 3, and the other 
lectures for the term, go to this question for stories in conversation in a fashion 
that yields analytic leverage on the notion “story” for students of stories-in- 
conversation distinct from stories in other contexts. 

Lectures 4 and 5 present, respectively, an extraordinary discussion of 
“entitlement to experience” (and to just the experience the events in question 
will sustain) as well as of the cultural organization of experience and the 
emotions, and a beautifully wrought account of “first” and “second” stories. 
But what I would like to call special attention to is the way in which Sacks 
brings an orientation to classical issues in social theory to a hypothetical - 
but compellingly plausible - account of cognitive organization (lecture 5, 
pp. 257-60). 

Using the metaphor of “designing minds,” Sacks asks how the preserva- 
tion of “experiences” might be organized. One cogent possibility might be to 
store experiences by what would commonsensically be considered their most 
important or salient aspects, or their most central character(s), or events, etc. 
As an alternative he proposes the possibility that experiences be stored “in 
terms of your place in them, without regard to whether you had an utterly 
trivial or secondary or central place in them” (p. 258). The consequences 
which this might engender - both for the organization of memory for 
experience and for social intercourse about experience - are then cast in terms 
of the concerns of social theory about the relationship between private interest 
and the public good. 19 

And that might have the virtue of providing a generalized motivation 
for storing experiences. If it’s your part in it that you use to preserve it 
by, then it might lead you to preserve lots of them, simply in terms of 
the idea of experiences being treatable as your private property. People 
can then collect a mass of private experiences that they then, by virtue 
of their generalized orientation to ‘what’s mine,’ have an interest in 
keeping. You might, then, design a collection of minds, each one storing 

19 A theme which Sacks had invoked as well in the account of turn-taking presented in Fall, 
1968. Lecture 4, pp. 50-4 and cf. discussion above, at pp. xiii-xiv. 



experience which is to be used for each others’ benefit, though you 
couldn’t necessarily say “Remember all these things so that you might 
tell them to somebody else.’’ You have to have some basis for each 
person storing some collection of stuff via some interest like ‘their own’ 
interest. Where, then, you get them to store experiences in terms of their 
involvement, but have them be available to anybody who taps them 

This sort of linkage between social organization and the organization of 
personal experience and its cognitive and emotional substrate - between the 
social, the psychological and the biological - will surely have to be successfully 
made eventually, and this is a novel and provocative direction in which it 
might be pursued. 

In its more immediate context, however, Sacks relates it stunningly to such 
diverse ancillary themes as the personal experience of being understood or not 
and the training requirements for professional therapists. 


Whereas the lectures for Spring 1970 were thematically coherent and 
focussed on storytelling, the materials for Winter 1971 (very likely a graduate 
seminar, rather than an undergraduate lecture course) deal with a congeries of 
more loosely related matters. But the central preoccupation is with “word 
selection’’ (cf. Sacks’ reference to “procedures whereby the words that people 
use come to be selected;” March 4, p. 308), and in particular those 
considerations of word selection that are often associated with “poetics. ” This 
set of presentations (complemented by the lecture for May 17 in the Spring 
1971 set, a lecture which deals with an eerie spatialization of metaphors, 
idioms, and other aspects of the talk of both parties in an emergency 
“psychiatric” phone call) constitute the basic point of depature in Sacks’ 
teaching oeuvre for this still largely unexplored domain of phenomena. 

This central preoccupation aside, special attention may be called to the 
presentation of March 1 1 which (at pp. 325-3 1) offers another one of Sacks’ 
astonishing tours deforce of analysis and interpretation. He starts with the text 
of a sequence which seems to be ordinary enough, even if in it a couple appear 
to press an offer of herring to an almost absurd extent. What Sacks does is to 
lay bare layer after layer of organization and preoccupation (on the parti- 
cipants’ parts) - from the differing grounds for making an offer than for 
re-making it, to the tacit relationships between the parties that emerge into 
relevance over the course of the sequence and come eventually to drive it, to 
the ways in which processes such as those which this sequence embodies can 
be a major component in both the stereotype and the enforced actuality of the 
elderly in a society such as this - that is, the United States in 1971. We cannot 
know whether the account which Sacks develops is biographically accurate for 
this particular family, but it feels compellingly on target for the sorts of 

Introduction xxvii 

interactional processes which can constitute the lived interactional reality for 
many persons. It is a signal display of Sacks’ ability to use a fragment of 
interaction to capture in an analytically compelling way a whole complex of 
social reality, from its social-organizational sources to its interactional embod- 
iment to its experiential consequences. This discussion presents as well both 
ends of a range of types of analysis which often appeal differentially to readers 
of conversation-analytic work. 

One end of the spectrum takes a particular episode as its virtually exclusive 
focus, with its scope of generalization being defined by “however this analysis 
turns out. ’ ’ Various particulars of context are traced through the full array of 
their consequences; here, for example, that the offer-recipient - Max - is a 
recent widower, and the offer-makers find themselves (on Sacks’ account) 
newly responsible for his well-being. The contingencies of the offer and its 
rejection, the relevance of pressing the offer and the import of its further 
rejections - all are understood by reference to these attributes of the 
participants, and the growing relevance of these attributes over the develop- 
mental course of the sequence. The account thus appears compellingly 

The contrasting end of the analytic spectrum focusses on the type of 
sequence involved, across variations in particular settings of enactment. For 
example, how is this sequence type — e.g., offers — related to other sequence 
types? Sacks had a long-term ongoing inquiry on request and offer sequences, 
and their relationship to each other and to other sequence types. Some of 
Sacks’ students have also pursued these questions in this more categorical 
fashion. Davidson (1984), for example, writes about “subsequent versions of 
invitations, offers, requests and proposals dealing with potential or actual 
rejection” (and see also Davidson, 1990). For dealing with the episode in the 
March 1 1 session Sacks finds it more in point to juxtapose 

an ‘offer’ as something different than a ‘request’ or a ‘warning’ or a 
‘threat.’ But in some situations the offer is simply the first version of 
getting the person to do something. 

That is, the mode of analysis being pursued can lead to different sets of 
alternatives providing the relevant comparisons or contrasts, “offer” making 
such alternatives as “warning” and “threat” potentially relevant here, even if 
they are not in other contexts. 

Though there may appear to be a tension between these two modes of 
proceeding, with the former often appearing more “humanistic,” “context- 
sensitive,” and “holistic” and the latter appearing more “formalistic” and 
“scientistic,” Sacks pursued them both. And the Winter 1971 materials show 
them pursued hand in hand - the word-selectional considerations being 
pressed in a generalized cross-context fashion, with this extraordinary single 
case analysis occurring in the same class session. 

xxviii Introduction 


In contrast to the Spring 1970 lectures which developed a coherently focussed 
account of storytelling in conversation and did so by sustained examination of 
a few data fragments, the Spring 1971 lectures (like those of the preceding 
term) vary both in topical focus (from sound ordering to professional-client 
interaction) and in data sources (from the “group therapy session’’ to a 
telephone call invitation from one student to another to a call to a suicide 
prevention center). 

Among these lectures are included a series - those for April 30, May 3, 
May 10, May 17, and May 21 - in which Sacks takes up the empirical 
materials which he addressed at the start of the 1964-5 lectures. Readers 
interested in the developmental course of Sacks’ lectures may wish to 
juxtapose these two treatments, separated by some seven years of intense 
intellectual work. It is not only that the same data are involved which might 
inform such a juxtaposition, but that themes reappear in the Spring 1971 
lectures which have not come up in the preceding several years. To cite but 
one example, there is Sacks’ discussion (May 21, pp. 405ff.) of the 
characterization of someone as a “stranger,’’ a discussion which goes back to 
the issue of categorization (though not in that technical terminology) taken up 
in detail in the paper “The search for help: no one to turn to” which was being 
written just before and during the first of these academic terms of lectures. 

Although the first lecture as delivered did include some initial discussion of 
a data fragment, 20 it was largely given over to the stance which Sacks was 
taking up with respect to his audience - both those present in the room and 
those interested from afar (including, therefore, the present readership). It is 
a rather franker statement than most instructors would give of the auspices 
under which they address an undergraduate class. And it reverses the 
relationship which might have been assumed to hold between the students 
sitting in the room and those far away - in place or time - who might be 
interested in “the work.” Rather than the latter being incidental and 
“by-product” recipients of materials designed for the undergraduates, it is the 
undergraduates who are recast as almost incidental onlookers to, and 
overhearers of, this analytic undertaking. 

“Almost;” for there is evidence throughout these lectures that the relevance 
of the co-present audience did in fact enter into the shaping of the issues and 
the manner of their presentation. There is, for example, the initial substantive 
discussion. 21 Sacks explores some ways in which speakers find or select words 

20 Cf. April 5 lecture, n. 1. 

2 'As in the case of several previous consecutive terms of teaching, Sacks begins the 
substance of the lecture set in the second of the consecutive terms with what he was exploring 
less systematically in the preceding term. (See, for example, Winter and Spring 1970 on 
storytelling in conversation.) 

Note that parts of the text here have been rearranged for the sake of continuity and 
coherence, so that some of the material included here with the lecture of April 5 was actually 
part of the introductory lecture. 



for use. In particular, he focusses on their doing so in “a history-sensitive” 
manner, for example, by reference to the sound or (later on, in the lecture for 
May 17) the metaphor composition of the prior talk. The tenor of the 
discussion is instructive. 

There are aspects of this discussion which suggest that, the stance taken in 
the first lecture notwithstanding, Sacks did not entirely ignore the nature of 
his co-present audience. The upshot which he takes from the discussion of 
sound-patterning (lecture for April 5, pp. 341-4) is (p. 343): 

for now . . . just to get some idea of how closely attentive in some 
fashion people are to each other, where picking up the sounds, doing 
simple contrasts, etc., are ways that they may be doing being attentive 
to each other. 

And again (p. 344): 

when we begin to collect the sorts of things that I’m noting here, we can 
feel that a serious attention to the way the talk is put together might 
pay. These sorts of things at least suggest some sort of close develop- 

And again, at the end of that lecture (in the present edition), after a discussion 
of strategic considerations relevant to the parties in the talk in the group 
therapy session materials (p. 347): 

And that paralleling of the attention to a distinctive weakness can 
suggest that they are moving with a kind of close attention to each other 
in a conflictive way. 

Two things may be said about the drawing of such conclusions. On the one 
hand, they are in point for hearers with no previous exposure to conversational 
materials and to this kind of close analysis of them. They seek to warrant the 
kind of attention being paid to these materials in a way that would not appear 
to be directed to an audience interested from afar in what Sacks has to say. On 
the other hand, it was a task to which Sacks recurrently addressed himself - 
to warrant these materials as respectable objects of study, and to establish over 
and over again, in a variety of respects, that these materials were orderly at 
quite refined levels of organizational detail. It is as if he were forever justifying 

— to others and to himself — the undertaking, its starting point, and its key 
premises. The upshots drawn here, early in the Spring term of 1971, can then 
be understood to be addressed not only, or not especially, to the students in 
the room, but to any recipients of his discussion. 

There are two matters taken up in the Spring 1971 lectures which have a 
history, either prior or subsequent, which it may be useful to call to attention 

— the relevant identities of conversational participants, and the notion of 

xxx Introduction 

In the lecture for April 19, Sacks begins a discussion of “caller-called” by 
reference to the possibility that 

some part of a sequential organization of conversation has to do with 
identities that the conversation itself makes relevant , such that for at least 
those facets of the conversation one needn’t make reference to other sorts 
of identities that parties have which are, so to speak, exterior to not 
simply the conversation, but to its sequential organization. If, however, 
we found that such other identities were central to almost anything one 
could say about a conversation, then there would be a way in which 
conversation could not be said to have an organization independent 
from such other aspects of the world as yielded other identities, e.g.; the 
names, sexes, social statuses, etc., of the parties. You could imagine a 
world where some social status the parties had, operated in such a way 
as to determine how 7 they could talk to each other, and in that world 
conversation would not be an independently organized phenomenon. 

The issue of the relevant formulation of the identities of participants comes up 
recurrently throughout Sacks lectures. In the Spring 1967 lectures, it may be 
recalled, there was a discussion of the possible “omnirelevance” of the 
category-set “therapist-patient” for the group therapy session materials - 
those categories straddling the line between “exterior” and conversation- 

Elsewhere, in the lectures of Fall 1967 (and even more centrally in early 
lectures for Spring 1967, not included in this edition) Sacks launched a 
discussion of turn-taking by considering a claim in a paper by the anthropol- 
ogist Ethel Albert (1964) about the practices of the Burundi. In this account, 
members of that society are all hierarchically ordered, and the society is small 
enough that on any occasion everyone present can assess their place relative to 
everyone else present. The distribution of opportunities to talk is organized by 
reference to this hierarchical ordering, 22 with the highest ranking person 
speaking first, then the next, etc., until each has had an opportunity to speak 
in an initial round; subsequent rounds reproduce this ordering. 

Leaving aside a variety of problems which can be expected in a system 
which worked in this way (and problems with the description), this account 
embodies what Sacks has in mind by “a world where some social status the 
parties had operated in such a way as to determine how they could talk to 
each other, and in that world conversation would not be an independently 
organized phenomenon.” 

The point is that, if one could show for some culture / society that there is an 
order or domain of conversation which is relatively autonomous of interac- 
tionally extrinsic attributes, then the possibility of such a culture would have 
been shown. Although it might be claimed in principle that there were other 

22 As Albert put it (pp. 40-1), “The order in which individuals speak in a group is strictly 
determined by seniority of rank.” 



cultures where there was no such autonomous form, there would then be 
certain burdens and opportunities of demonstration and exploration that 
would have to be addressed. For one, it would have to be shown that there 
was an empirical instance of such a society/culture, rather than simply be 
asserted in the nature of the case. For another, it would be a feature of a 
society/culture which might then be explored for what else it was related to, 
how it came to be so, how it was embodied or implemented, etc. 

This discussion, then, is intimately related to the issue of the omni- 
relevance of gender or class status, etc. (class and gender are singled out here 
because they are the features most often invoked as specially constraining and 
shaping the conduct of talk in interaction). For to show a relatively 
autonomous order of organization (or several such orders) for conversation 
would be to establish domains of interaction not necessarily contingent on 
gender, class, etc., and thereby to show conversation to be “an independently 
organized phenomenon.” 

One significance of the categorical identities “caller-called” is that they are 
conversation-specific (unequivocally so, unlike “therapist-patient”), and it 
appears that they serve as the feature by reference to which various aspects of 
talk are organized, especially with respect to the overall structural organiza- 
tion of single conversations. This had been shown in Schegloff (1967, 1968) 
for the organization of openings, e.g., with respect to who talks first. Part of 
Sacks’ argument here turns on the relevance of caller/called not only for 
openings, but for “closings,” for example, it being the caller’s business to 
initiate arrangement-making and other ways of getting to the end. 

It is striking that in a prior discussion of omni-relevance (in Volume 1, 
Spring 1967, lecture 14, and cf. the introduction to Volume 1, pp. liii-liv), 
Sacks argued for the omni-relevance of “therapist-patient” in the group 
therapy sessions by reference to its being the therapist’s business — in that 
capacity - to bring the session to a close, and that a new patient has to be told 
that an “indirect” closing initiation by the therapist was doing that job, 
something which he did not himself see and which it would not have been 
doing had anyone else said it. The relationship of some identity to a bearing 
on “closing” (at least of a conversation as a whole) may, then, turn out to be 
of strategic importance in showing category omni-relevance. 

The issue of the relative autonomy of conversation/interaction has had a 
continuing relevance for students of interaction. Perhaps the most prominent 
discussion of the issue within contemporary sociology was Erving Goffman’s 
presidential address to the American Sociological Association, ‘The interaction 
order’ (Goffman, 1983) which also argued (albeit along different lines) for the 
relative autonomy of the organization of interaction from other aspects of 
social organization. 

Another topic with a considerable later development figures in these 
lectures for Spring 1971. At the end of the lecture for May 24 (pp. 414-15) 
there is a discussion (the first of which there is a record, though Sacks refers 
to an earlier related lecture) of the asymmetry of “yes” and “no” answers - 
related to the form which a preceding “yes/no” question has taken. This is 

xxxii Introduction 

an early form of what would eventually become, under the name “the 
preference for agreement” (cf. Sacks, 1987 [1973]), a much more general 
account. Here Sacks appears to focus on such questions as might be termed 
“pre”s, even if the future course of talk - the type of sequence - which the 
speaker meant to undertake is not at all clearly projected. Whereas the 
discussion in the May 24 lecture is quite specifically situated, and refers to 
courses of action in which some sort of sequence may figure as preparation or 
“setup,” Sacks would two years later, in the public lecture at the Linguistic 
institute in Ann Arbor on which the 1987 publication is based, depict the 
preference for agreement as a much more general - structural - feature of 
question/answer sequences of the “yes/no” type, with still more general 
implications, for example, for adjacency pairs. 

A key component of this notion is that of “preference,” and it has a longer 
(and variably focussed) history in Sacks’ oeuvre. In these lectures, for example, 
in the lecture for April 23, Sacks proposes that some formulations of the event 
for which an invitation is being tendered are “preferred:” if the occasion is to 
include dinner, for instance, the invitation should be for “dinner;” anything 
else (e.g., “drinks”) and recipients will hear that it is “not for dinner,” for, 
given its “preferred” status as an invitation form, it would have been used if 
it could have been used. And two years later at the Linguistic Institute, where 
the “sequentialized” version of “preference” was extended from the usage 
here to that of ‘On the preferences for agreement and contiguity. . . the 
application of the notion “preference” to “formulations” was extended from 
formulations-of-events to reference-to-persons in the drafting of the paper 
‘Two preferences in the organization of reference to persons in conversation 
and their interaction’ (eventually published as Sacks and Schegloff, 1979). 

And before the usages here in the Spring 1971 lectures, a similar notion 
underlies the conception of ‘specific alternatives” (cf. Schegloff and Sacks, 
1973: 305, 313-14; this paper was first drafted during the summer of 
1969). There the notion of “specific alternatives” made relevant by an 
utterance such as a “possible pre-closing” was explicated by noting (p. 314) 

the alternatives made relevant by an utterance of that form are not 

symmetrical. Closing is the central possibility, further talk is alternative 

to it; the reverse is not the case. . . 

That feature of asymmetry - later central to the notion “preference” - came 
up in other working sessions of 1969 and 1970. For example, I recall Sacks 
remarking on it while examining tapes made by Melvin Pollner in a Southern 
California traffic court; the observations concerned the treatment by the 
parties involved of the source of income of a college student appearing before 
the judge; Sacks took it that some sorts of financial support (I do not now 
recall which) were central and “normative” (in the sociological terms of the 
time), and others were alternatives to them, but not vice versa. 



The contrast figures, in essentially these terms, in lecture 6 for Fall 1971, 
at pp. 455-6, where Sacks is discussing a story told by a teenage girl to a 
teenage boy, a story which turns centrally on her spending half the night with 
a “guy that {she} liked a real lot.” What is central to the telling of the story 
is that they spent the time “in the back house” (i.e., the house behind the 
main house, a sort of guest house) instead of “in a car.” Sacks shows how “in 
the car” is built into the story as “normal” for teenagers, something with 
which the teller is trying to fashion a contrast. “In the back house” is then a 
specific alternative; it is an alternative to “parking” or “in the car,” but the 
latter is not “an alternative;” it is the basic, unmarked (as linguists might put 
it) place. And in that same context Sacks introduces the use of the term 
‘ ‘preference: 

. . .She can. . .invoke the normal priorities, in which, for unmarried 
teenagers, parking is ‘preferred.’ I don’t mean that it’s favorite, but 
there’s some way it’s preferred over the back house, if at least only in 
moral terms. That is to say, she brings off that she prefers the back 
house, but there is a more abstract sense of ‘prefer’ which involves her in 
invoking the parking — that which is ‘preferred’ in the more abstract sense 
— as a first alternative” {final emphasis supplied} 

It is this sense of “preference,” as “a first alternative, to which others may 
contrast but which itself does not contrast with them” which is one central 
thrust of subsequent uses of the term, both by Sacks and by most others 23 


Although the particular phenomena and data sources taken up in the lectures 
for Fall 1971 are quite different, the thematic commitment underlying this 
course is strikingly reminiscent of the lectures for Spring 1966. In both may 
be found explorations of how (a) culture is to be conceived which blend a 
fresh theoretical conception with a distinctive and organic relationship to 
“ordinary” conversational data. 

One relevant bit of background for the first lecture of the term may well 
be an episode in law school (earlier recounted in the introduction to Volume 
1, n. 6) which alerted Sacks to the mysteries of commonsense assessments of 
the plausibility and seriousness of conjectured events. Law students debating 
a point in the law of torts rejected as implausible the premise of an airplane 
flying at an altitude of five feet while willingly discussing hypotheticals only 

23 This includes, for example, Schegloff, Jefferson and Sacks, 1977. The notion captured by 
“preference” figured in my own work at the time as well (especially in Schegloff, 1970), but 
not under any of these names. For further discussion, cf. Schegloff, 1988, and, for another 
view, Bilmes, 1988. For applications, discussions and reviews of the notion of “preference” 
(and “dispreference”) cf. among others Atkinson and Drew, 1979: chapter 2; Heritage, 
1984: 265-80; Levinson, 1983: 332-45; Pomerantz, 1978, 1984. 



marginally larger. The concerns then awakened, which had driven the 
insistence on actual observation and then recorded and re-examinable data, 
may have served tacitly as the grounds for the line developed at the start of 
lecture 1 on the trouble of dealing with imagined occurences, and the 
impossibility of dealing with events that strain commonsense credibility, 
events which otherwise can be shown to be real. 

This theme reappears explicitly later in the Fall 1971 lectures, in Lectures 
9-12 “On some technical considerations of a dirty joke.’’ 24 There Sacks 
points to a staggering number of implausible co-occurances on which the 
story/dirty joke being examined depends, and without which it collapses. 
And he addresses himself especially in lecture 10 to the devices by which the 
telling of the joke/story can survive these apparent implausibilities. He rejects 
an “Aristotelian” solution along the lines of a generic “suspension of 
disbelief’ by noting that no disbelief arises to be suspended, and that the story 
could not survive if it did. He suggests instead that a recipient is fully engaged 
in understanding the story, and that the artfulness of the story in deploying 
the elements from which an understanding can be achieved channels attention 
in a fashion which circumvents the implausibility by naturalizing and 
sequentializing the events. 

Still, the isolation of this problem and the treatment of the narrative form 
by ironic comparison to a quasi-realistic story suggests a continuing underly- 
ing preoccupation on Sacks’ part with the relationship between the real and 
the unreal, the plausible and the implausible, the real and the plausible, the 
real and the implausible, etc. And here again (as in the Spring 1966 lectures; 
cf. introduction to Volume 1, pp. xxxix-xli) may be found echoes of the 
“commentator machine” introduced in Sacks’ early (1963) paper ‘Sociolog- 
ical description,’ with its metaphorically articulated depiction of various 
possible relationships between real doings and the accounts offered of them, 
and the account-offering as itself a real doing relative to which another doing 
may be a defective exemplar. 

Lecture 1 begins an announced preoccupation with “storytelling in 
conversation” with an observation about a pun, and the first several lectures 
are as much about puns and proverbs as they are about storytelling. 

Sacks’ concern with puns, which would eventually issue in a presentation 
at the Georgetown Roundtable in March, 1972 and the little paper (Sacks, 
1973) ‘On some puns, with some intimations,’ is analytically located at the 
intersection of problems of word-selection of a “poetics ’’-like character on the 
one hand, and the practices of storytelling sequences on the other. His 
discussion of puns here in lecture 1 as well as in the Georgetown Roundtable 
paper is focussed on their use 25 by a story-recipient just after story completion. 
The occurrence of puns - unintended and unheard puns - in this distinctive 
sequential position may have recommended itself to Sacks as a case in point 

24 These lectures were published under that name, as edited by Gail Jefferson, cf. 
Schenkein, 1978. 

‘ 5 And the use of proverbs; cf. the ensuing discussion below. 



for the contrast between “implausible ‘real things’ ” and imagined things one 
could get someone to believe as a basis for theorizing about them. Or perhaps 
the order was just the opposite; entertaining the possibility of opening the 
lecture and the course with a discussion on puns, some groundwork seemed 
called for, addressed discursively to the believability of the sort of thing he was 
going to begin with. 

The discussion of proverbs (at lecture 1, p. 422) goes back to Sacks’ 
reaction while still a graduate student at Berkeley to the beginning of George 
Homans’ book, Social Behavior: Its Elementary Forms (1961). There (pp. 
1-2) Homans remarks in passing on the traditional folk sociology which (it 
was apparently his view) it was scientific sociology’s business and mandate to 
correct and supplant. The hallmark of this folk knowledge was for Homans 
apotheosized by proverbs with their “obvious” truths, but also directly 
contradictory truths. 

This theme was by no means unique to Homans. Both Homans and the 
many others who contrasted scientific sociology with common knowledge 
were engaged in defending sociology as an academic discipline from charges 
that it was “nothing but common sense.” Many took it that one line of 
defense was to show the weaknesses of commonsense knowledge, and thus the 
proper office of sociology in reference to it. That office was to replace 
“common sense” with something more scientific. This was, of course, one 
central point of reference for Garfinkel’s observation (1967: chapter 1) that 
the social sciences were addressed endlessly to the substitution of “objective” 
for “indexical” assertions, and the alternative ethnomethodological program 
which he put forward - to make commonsense knowledge non-competitively 
a topic of sociological inquiry. 

Sacks was struck early on (that would have been in the very early 1960s, 
most likely in 1962-3, while we were at Berkeley’s Center for the Study of 
Law and Society; cf. introduction to Volume 1, p. xv) by Homans’ 
non-analytical, non-sociological stance toward proverbs - treating them as 
primitive and faulted versions of scientific propositions. The issue for Sacks 
was, precisely, what were proverbs (as natural objects, so to speak), and what 
were they used to do, that might make the features which Homans treated 
ironically seriously understandable. He sought out a relevant literature and 
found Archer Taylor’s The Proverb, but did not find the answer there, though 
he respected it as a work of scholarship. 

It is striking then to read Sacks’ treatment here (at lecture 1, p. 422) 
with this history in mind. Briefly, in the context of a discussion of the use of 
proverbs by story recipients on story completion, and having remarked on 
the common observation of the inconsistency between different proverbs, he 

Now the question is, is that a defect of proverbial expressions? Or is it 
that, if it turns out that what proverbial expressions do is that they are 
used to understand something else, then the question for them is, are 
they applied to something that they evince an understanding of? If so, 



it’s quite irrelevant that, as a package, they can turn out to have an 
inconsistency among them. The problem is not, on any given one’s use, 
is it true relative to other proverbial expressions, but, does it, as 
something one understands with, understand what it applies to? Where, 
what it applies to is the story it’s used after. . . 

. . .What’s done with them is to take one and see how, for what it’s 
positioned after, does it understand that. It can then be seen as 
irrelevant, somewhat arbitrary, to say “Let’s take the set of them and 
consider whether they’re consistent, to determine whether they’re true.” 
That may be not at all how, empirically, they work. 

Here, some ten years later, is Sacks’ answer to Homans - his contrasting 
account of how proverbs should be treated by sociologists. And in this little 
passage is the direct confrontation of the effort to treat proverbs as defective 
propositions - failures as “objective” expressions - with the claim that they 
are designed fundamentally as objects for indexical deployment. They are 
meant specifically to display understanding of the local object they are placed 
after - they are prototypically indexical in that sense. Each is to be juxtaposed 
to its occasion of use, for which it was employed; that specifically renders 
problematic the detachment of each from the environment for which it was 
produced, for juxtaposition with other such disengaged-from-context objects. 
And Homans’ critique of them — based on just such a disengagement — is the 
apotheosis of the social science practices to which Garfinkel meant to set 
ethnomethodology in “non-ironic” contrast. For Sacks, this analysis grew 
directly from his effort to figure out how proverbs worked. 


Those familiar with the published corpus of conversation-analytic work will 
recognize in lecture 5 a version of the ‘Two preferences. . .’ paper (Sacks and 
Schegloff, 1979). I do not recall at what point Sacks and I found ourselves 
both focussing on the contrast between what Sacks here terms “Type 1” and 
“Type 2” identifications of persons and what I was calling (with no restriction 
to the description of persons) “description-for-recognition” and “description- 
for-understanding.” The written version of the paper was initially drafted by 
Sacks while we were living in the same house during the Linguistic Institute 
at the University of Michigan in 1973. I did not know he was drafting it, 
until he gave it to me early one evening to look over. Although we worked 
over it intermittently, the changes made from the initial draft were relatively 
small and technical. 

The discussion of forms of reference in lecture 5 (as well as the paper which 
followed) can be located in another course of development. While still at 
UCLA (probably about 1966) Sacks had drafted a paper which came to be 
referred to as the “two-person identification” paper. The data fragment which 
had given rise to the “two-person identification” line of analysis, and around 



which the paper was written, was taken from the observational materials 
collected by Barker and Wright. 26 In this episode, a little girl enters the 
kitchen of her home and finds her mother talking to another woman, 
someone the little girl does not know. The following exchange is then 

Little girl : Who is she? 

Mother : That’s Rita. Do you remember the other day when you went 

to the party and met Una? Well that’s Una’s mother. 

This data fragment was appreciated against the following analytical back- 

Sacks had established in his dissertation work (cf. the published version in 
Sacks, 1972a) that there was no general solution to what he termed “the 
one-person identification problem.’’ That is, faced with the task of 
identifying/categorizing a single person, there were demonstrably available 
multiple “membership categorization devices” which contained some cate- 
gory which could properly categorize any person. 27 And there did not appear 
to be any general solution for selecting which device to use - no general 
preference rule that would select some device from among whose categories 
“the correct one” for the person being categorized should be selected. This 
was a finding with many analytic and theoretic reverberations. For example, 
analytically, any actually employed categorization employed by a speaker in 
talk-in-interation had then to be viewed as a contingent product whose 
achievement could be subjected to analysis by reference to the particulars of 
its local environment. (And, theoretically, social scientists’ categorizations 
could not be warranted solely on the basis of their descriptive correctness, but 
had to be otherwise warranted, e.g., by reference to their relevance, whatever 
grounds of relevance might be chosen.) 

What the data fragment reproduced above suggested to Sacks was that, 
whereas there seemed to be no general solution to the owe-person indentifi- 
cation problem, there might be a solution to a two - person identification 
problem. In his dissertation (1972a), he had described what he proposed to 
be a categorization device composed of pairs of linked terms - “paired 
relational categories” he called them - (e.g., friend-friend, husband-wife, 
relative-relative, parent-child, neighbor-neighbor, stranger-stranger, etc.), 
which constituted “. . .a locus for a set of rights and obligations. . .” (p. 37). 
This categorization device was used to categorize a population of persons not 
one at a time, but two at a time — as incumbents of one of these paired 

26 Although Sacks had worked on some observational materials which Barker and Wright 
had published (for example, One Boy's Day , 1951), I believe the fragment involved in the 
“two-person identification” paper was taken from other, unpublished, material of theirs 
which Sacks had secured. 

27 Sacks had termed these devices “Pn-adequate,” i.e., adequate for any, unspecified (hence 
“n”) population (hence “P”). The devices/collections of “age” and “sex/gender” categories 
were his most commonly invoked instances. 

xxxviii Introduction 

relational categories. It appeared as well that there was only one such 
categorization device, only one which identified / categorized persons two at a 

What the “Rita/Una’s mother” data fragment suggested was that one 
way members might handle a one-person identification problem which had 
no general solution was to transform it into a two-person identification 
problem which did have a general solution. In the instance at hand, the little 
girl’s mother adopts this solution: asked to identify one person (“Who is 
she?”), she introduces another person into the identification problem - Una - 
and then identifies the pair of persons by a set of paired relational categories: 
mother-fchild] (“. . .That’s Una’s mother”). This was an extremely elegant 
solution to the identification problem, and an extremely elegant account of it. 

But there were problems, and on a visit to the west coast during the winter 
break Sacks and I discussed them at length, as we regularly did with one 
another’s written work. The most telling - and ultimately fatal - problem 
was that this solution did not work as a general solution. For one thing, not 
all the paired sets of terms could be (or were actually) used by interactants; for 
example, although “stranger-stranger” was one of the paired term-sets (and 
one indispensable for the empirical context which first gave rise to the 
formulation of this categorization device), persons confronted with an 
identification problem do not respond by saying, “That’s Rita. Remember 
Una? Una and Rita are strangers.” Were stranger-stranger an eligible 
category-set for these purposes, there might be a general solution to the 
one-person problem by converting it into a two-person problem. Without it, 
it was not a general solution. 

Another problem, equally fatal and with clear connections to the lecture 
which prompts this discussion (and to the ‘Two preferences. . .’ paper), was 
that not any person could be introduced as the second for co-categorization 
with the initial person to be identified, and not even any person in a specified 
range of relationships to the target problem. Only such persons could be 
introduced (or seemed actually to be introduced) as were expectably recog- 
nizable to the one posing the problem, the one for whom the categorization 
was being done. So again, persons confronted with an identification problem 
do not say “That’s Rita; there’s a little girl named Una, and Rita is Una’s 
mother.” There was, then, not only a constraint on which set of paired terms 
could be used for the target person and the one to be introduced as second; 
there were restrictions on which second person could be introduced for this 
purpose by reference to the knowledge of the recipient of the identification. 
Indeed, the possibility could not be ruled out that no second person could be 
found who would satisfy both containts (nor was it clear that these were the 
only constraints). The status of this categorization device as a general 
solution to a two-person identification problem was thus cast into doubt, let 
alone its status as second-order solution to the one-person identification 
problem. The “two-person identification” paper was shelved. But it was 
not without consequences, of which brief mention can be made here of only 



First, the data fragment which motivated that earlier effort has here, in the 
Fall 1971 lectures (lecture 5, pp. 451-2) become an example of 

where a speaker doesn’t figure that recipient knows who’s being referred 
to, but knows something that involves it in being an ‘almost,’ i.e., that 
you know someone in some close relationship to that one being referred 

Its bearing thus is incorporated into the discussion of “recognitional refer- 
ence” and the preference for recognitional reference even when the possibility 
of its achievement is open to question (cf. Sacks and Schegloff, 1979). 

Second, it seems to be relevant in a curious way to a tack taken in an earlier 
lecture, and on quite a different topic - lecture 6 for Fall 1968. A bit into that 
lecture (at p. 70), Sacks is discussing introductions (of one person to another), 
what occasions them and how they’re done. 

One way to think about it is to consider that a way to simplify the task 
of doing any introduction would be, e.g., to constrain the occasions 
under which introductions could get done. You could say, for example, 
introductions should go ‘first name to first name.’ That can operate to 
constrain the initial use of an introduction to only people you can 
introduce that way. 

But, Sacks points out, that runs up against the fact that the conversations 
within which introductions have to get made are generated by an entirely 
separate mechanism from the one that makes introductions possibly rele- 

The relationship to the problems with the “two-person identification” 
paper is this: one problem with that paper, as just recounted, was that the 
mechanism only worked for certain possible values of paired relational 
category terms (not for e.g., stranger-stranger), and setting such a pre- 
specification subverted the potential generality of the device. So here as well, 
where the point is that an introduction mechanism is needed which will have 
as general a scope as whatever occasions the relevance of an introduction and 
whatever occasions the already-ongoing conversation within which introduc- 
tions come to be relevant. Pre-constraining introductions to certain values of 
introduction terms would subvert the viability of that institution. This is just 
another specification of the more general result that pre-constraining the 
elements of a device which can be employed subverts the possible use of the 
device as a general solution to some problem in the practices of interaction. 

Third, the working through of the problems of the “two-person identifi- 
cation” paper seems to have deeply affected Sacks’ thinking about the relative 
merits of single case analysis versus the use of aggregates of data for the 
purposes of building a discipline. Note that the issue is not the status of single 
case analysis per se, but the possibility of building the sort of desired discipline 
which had come to be the goal of conversation-analytic work. In a letter to me 



a few years later (March, 1974), Sacks remarked on the relevance of 
“working with masses of data” as what “in the end differentiates what we do 
from e.g. French structuralism.” And, in this regard, he invoked the 
experience of the problems with the “two-person identification” paper - and 
its effort to ground a general solution in a single case - as evidence enough. 28 


Having initially projected the Fall 1971 lectures to be about stories and 
storytelling, the first six lectures depart somewhat from a close focus on that 
topic, although remaining at least tangentially relevant with lectures 7 and 8, 
and then with the series of lectures from 9 through 12, Sacks comes back 
squarely to his announced topic. 

Lectures 7 and 8 address the “motive power of stories.” The theme is a 
penetrating and remarkable account of a particular class of stories. These are 
stories which come to be retold after a long time delay (“long” here meaning 
years), a delay during which one who had been the recipient of the story 
becomes the kind of person the teller then was, and tells it in turn to a 
recipient such as he was when he was told it - the retelling being done on just 
the sort of occasion which is appropriately analyzed by the story. Such a 
“delayed-fuse” story thus serves as a kind of cultural repository for 
occasion-ally relevant knowledge. (The material being analyzed involves an 
older man, seeable as “no longer having prospects,” telling a younger man, 
who is about to depart for college - and prospects - about the time he was a 
young man, with prospects, and what became of them.) 

These lectures call to mind the lectures of Spring 1966, for the way in 
which they speak to the nature of culture, the ways in which culture mobilizes 
minds as a repository of what it has to transmit, and uses stories as the vehicle 
for transmitting that knowledge, recruiting the interactional stances of the 
participants in the situations in which they find themselves - for which the 
stories provide analyses - as the energy driving the telling of the stories as 
matters of e.g., self-justification. They also recall lecture 5 of Spring 1970 on 
how memory for experiences can be motivated by having them stored as “the 
property” of the one to remember them, to be accessed by others by telling 
a “similar” story. 

The theme plays off a by-now cliched geneticist “witticism” that chickens 
are the device by which eggs reproduce themselves. Here persons, their 
experiences, and the stored versions of experiences in stories are the device by 
which culture reproduces itself and adapts to changing social circumstances. 
The line taken here is reminiscent of a term (though not necessarily the 

28 He wrote, “The ‘structures for particulars’ direction [which is how Sacks had earlier 
characterized “the thrust of my stuff over the years”] doesn’t work: recall the two-person 
paper failures, etc. and the ‘system for masses,’ for routine, etc. may.” (The internal quotation 
marks have been added for clarity.) 



correlative meaning) which the anthropologist Alfred Kroeber introduced 
some years ago for culture - “the superorganic.” 29 

These resonances of lectures 7 and 8 are sustained in the following four 
lectures, 9 through 12, on the dirty joke as a technical object. In this 
discussion as well, the story form is treated as a packaging device for elements 
of culture, as was the case in lectures 7 and 8. There is a distinct shift in theme 
here, focussing less on the teller doing things via the story and more on the 
story doing things through the teller, and doing them through the teller as the 
instrument of a culture. The story in general, and the dirty joke as a technical 
object in particular, get worked up somewhat formally here in a fashion 
parallel to the account of games (and children’s games in particular) as 
packaging units for a culture in the Spring 1966 lectures. 

This is a weighty theme and it may be appropriate to understand Sacks to 
have prepared his audience for it in the opening lectures for the term. Recall 
that in the first lecture in this set for Spring 1971 Sacks had tried to provide 
grounds for taking seriously the possibility that there really was a pun in the 
story, that it was not just a “reading-in” by the analyst, just as he had done 
in other first lectures, to ground the seriousness of word-selectional or 
“poetics” observations. Here he proceeds in the same fashion by showing the 
“artfulness” of the dirty joke/story, the elaborate way it is put together in 
order to ground a claim for its status as a technical object, and eventually his 
claim for it as a serious transmitter of culture. 

It is in lecture 1 1 that the theme of the dirty joke as a packaging device for 
culture, with its “dirtyness’ serving as a form of restriction on its circulation, 
is stated most pointedly. It may be worth mentioning here again (cf. 
introduction to Volume 1, p. xxii) the relevance of the work of classical 
scholars such as Milman Parry, Albert Lord, and Eric Havelock, all of whose 
work Sacks was familiar with, and from whom he would have become 
familiar with the notion that the classic forms of oral cultures - such as the 
Greek epic - served as major instruments for the preservation and transmis- 
sion of a culture, the story line of the epic being not so much the point of it 
as the shaper and guarantor of its transmission. It was just one aspect of the 
special metier of Sacks’ mind and sensibility to see in this juvenile “dirty joke” 
told in a teenagers’ group therapy session the contemporary operation of so 
grand a theme, otherwise treated as the special preserve of elite “culture.” 

Another echo of the Spring 1966 lectures in Fall 1971 is the appearance 
of a concern with children, and children’s learning the ways of the culture and 
its rules, a theme which is central here in lecture 12. This lecture again calls 
to attention Sacks’ extraordinary capacity to take apparently general views 
and characterizations of the world, ones which present themselves as 
“natural” accounts of it, and to specify them, often showing them to embody 
some distinct and limited perspective. Thus in lecture 12 he depicts what 
seems to be a potentially anybody’s recounting of a scene as specifically 
embodying the perspective of 12-year-old girls. In the earlier lectures 7 and 

29 Cf. Kroeber, 1917. 



8 for Fall 1971, he shows how a story embodies in particular the perspective 
of persons with live prospects for a future or those with already failed 
prospects. And in an earlier set of lectures (Winter 1970, lecture 2) he takes 
what appears to be a passing observer’s “bland” general account of a scene in 
which “the police were handling some trouble at a department store” and 
shows that to other eyes - of members of different social groups with a 
different experience with authority - what might be seen might be not that 
there was trouble and the police came and handled it, but that the police came 
and there was trouble and it was unclear how it was being handled, how it 
would turn out, and how it would turn out to implicate them. 

Although seemingly quite remote from that tradition of analysis, these are 
exemplary exercises in the sociology of knowledge. Apparently unsituated 
views and understandings of the world and of particular settings — otherwise 
understandable as just “how things are/were” - are analyzed for the 
distinctive social groups to which they are affiliated, and with whose 
experience of the world they link up. These discussions illuminate our 
understanding both of the particular settings and utterances being addressed, 
and of the distinctive experience of social circles to which we gain access by 
way of these discursive practices. 

In this regard, it is striking that one of Sacks’ characteizations of the special 
perspective of 12-year-old girls by reference to which the dirty joke being 
examined in lectures 9-12 should be understood is reminiscent of his 
depiction of the perspective of suicidal persons who see themselves (and report 
themselves) as having “no one to turn to.” That phrase supplied the subtitle 
of Sacks’ major early paper (and his dissertation), The Search for Help: no one 
to turn to. Now remembered primarily for its formal statement of the 
categorization problem and aspects of its solution, it may be useful to recall 
that, although textually at the beginning of that work, developmentally it was 
subsequent to the initiating problem, which was how someone might come to 
say “I have no one to turn to,” and say it seriously (that is, as the reported 
result of a search procedure), delivered paradoxically precisely in an occasion 
in which it seems apparent that they have found “someone to turn to.” Sacks 
began with that, although in the paper he ends with it. 

The proximate solution of “no one to turn to,” Sacks proposed, was that 
the person involved (the suicidal person, that is) had such a problem as would 
alienate precisely the person(s) whom the normative search procedure would 
locate as the proper persons to turn to. That is, there are in general “persons 
to turn to” (formulated by reference to paired relational terms discussed 
above at pp. xxxvii), but the problem involved, if reported to those persons, 
might lead to their abandonment of just the status which made them the 
one(s) to turn to. Thus, for example, turning to a spouse with a problem 
engendered by one’s adulterous involvement. 

What is striking is the formal similarity to this situation of the putative 
circumstances of 12-year-old girls in Sacks’ account of the dirty joke: namely, 
the problem of checking out information about sex, information acquired 
illicitly, e.g., by listening in to the parent’s bedroom from behind a door: with 



whom can such supposition be checked? the parents one spied on? the friends 
to whom one cannot reveal just this inexperience? The formal similarity is 
striking: the nature of the problem is what precludes turning to just the ones 
one would otherwise turn to for its solution. 

The last lecture for Fall 1971 is about dreams, and seems quite disengaged 
from the other lectures. In fact, Sacks developed a considerable interest in 
dreams (among other respects as a format in which stories are preserved), and 
pursued it, largely informally, during these years. In part, this had developed 
from his reading in Freud and in a variety of literary sources; in part it 
converged with an interest in popular culture (an interest which, in the 
last several years before his death, included such matters as advertising as 
well.) He was, for example, interested in the presentational modality of 
dreams - whether they were experienced as being read or being seen in action; 
if seen, like a movie, whether they were in color or black and white; what 
sorts of editing and directorial techniques informed their structure, and the 


The lectures for Spring 1972 begin in the same fashion as did those for Fall 
1968, as a systematic and general account of an organization or a class of 
conversational occurrences - in this case, adjacency pairs and adjacency pair 
organization. It is not until the second half of the second lecture that a 
determinate, actual (as compared to intendedly exemplary or “characteristic”) 
bit of talk is presented for careful and detailed examination. But the text of 
the first lecture and a half nonetheless makes clear that this general and 
systematic introduction to the projected subject-matter for the course is based 
in a detailed way on a substantial corpus of observations and analyses of 
particular stretches of talk of which adjacency pair organization is to be offered 
as a tentative account (though hardly preparing us for the illuminating detail 
exposed when the first bit of data is examined closely in the second half of the 
second lecture). 

The general features of adjacency pairs are first introduced via a variety of 
particular sequence types - greetings, terminal exchanges, question-answer 
sequences, etc. - each of which names its own, recognizable class of sequences. 
Adjacency pairs are thus introduced as a class of classes. But the particular 
variety of sequence types is strategically selected to display something of the 
extraordinary provenance of adjacency pairs - used at the critical junctures of 
virtually all the main kinds of organization of conversation: at the opening 
and closing boundaries of particular episodes of conversation, as the central 
device by which next speakers are selected, as the basic tool for remedying 
various locally occuring problems in conversation, as the locus for departures 
from a single-sentence format for utterances (sub-sentential utterances char- 
acteristically being second-pair parts, and the construction of multi-sentence 

xliv Introduction 

utterances being mediated on this account by adjacency pair constructions), 
and so forth. 

This introductory account of the generality of adjacency pairs by reference 
to the other types of organization in which they figure prominently and 
strategically is followed by another, an account of their provenance by 
reference to their distributional generality. That is, if we ask where adjacency 
pairs can go (and, in particular, where their first pair parts can go, since where 
the second pair parts go is given by the first pair parts, i.e., after them), we 
find that their privilege of occurence is unrestricted except by reference to 
adjacency pair organization. That is, they can go anywhere except after a first 
pair part, unless the one going “after” is initiating an “insertion sequence.” 
The point here is two-fold: our sense of the centrality of adjacency pair 
organization is reinforced by its virtually unrestricted distribution, and our 
sense of its basicness is reinforced by its self-organizing character, that is, by 
the observation that the only restriction on its distribution is that imposed by 
adjacency pair organization itself. (Recall that a similar argument had been 
offered for the basicness of the turn-taking organization in Fall 1968, lecture 
4, pp. 54-5, and this introduction, above, p. xvi and n. 12). 

When Sacks turns to the examination of a specimen of an adjacency pair, 
the focus shifts sharply. The exchange - a question/answer sequence - 
occurs in 

a telephone conversation between two middle-aged women one of 
whom has gone back to college part-time, and is telling the other about 
a class she’s taking 

The other - Emma - asks: 

Emma : Are you the oldest one in the class? 

Bernice : Oh, by far. 

In some five pages, Sacks shows an array of issues to be involved which most 
readers, I suspect, will not have anticipated. Here I want to draw out one of 
them, one which echoes themes raised in earlier lecture sets, especially that for 
Spring 1966 (and see the introduction to Volume 1, pp. xxxvii-xxxix). The 
issue concerns the proper understanding of the positioning of the subject- 
matter of these lectures - and of the area of inquiry which has developed with 
the name “conversation analysis” - among the disciplines. 

One of Sacks’ early observations about this exchange is that the question is 
not characterizing Bernice’s position in the class as one of a possible set of 
positions, others of which might be “second oldest,” or “one of the oldest,” 
and the answer is not just a way of saying “yes,” or saying it emphatically. 
Rather, Sacks proposes with respect to the former, the question is asking 
about a “unique position” in the class, with a variety of features which can go 
with occupying a unique position (“being the only X”); in that respect, its 
relevant alternatives are not the set of age-grade positions, but things like 

Introduction xlv 

“Are you the only woman? Disabled person? African American?” etc. Sacks 
continues (lecture 2, p. 538), 

So that what seems like a kind of obvious semantics turns out to be 
wrong for our language. It’s one you hear around, and it says: Take “the 
oldest one in the class” and find its meaning by considering the set of 
alternatives to it, where the alternatives can easily be derived from it by 
just considering some obvious way in which it is part of a set of positions 
having to do with ‘oldness.’ . . . Now, alternatives are an obvious way 
to go about locating what something is doing or what something means. 
But the question of alternatives does not have an easy answer. It is, for 
any given thing, an empirical issue and not simply a transparent 
semantic issue to be gotten by lexical considerations. In saying what I 
figure to be the kinds of things that are alternatives here, both in the 
question and in the answer, I’m saying something that has to he 
discovered from a consideration of the way the world works that produces 
these kinds of sequences. This obviously produces a massively complex 
set of problems in analyzing things like a small question-answer 
sequence. For each one of them, if we’re going to use alternatives to find 
out what it means, then we’re going to have to go into a discovery of 
what the alternatives are. [Second emphasis supplied} 

The point to which I wish to call attention is that this is not a matter of 
linguistic analysis in the usual sense; the closest might be some form of 
anthropological linguistics or linguistic anthropology, though those disciplines 
have shown qualified enthusiasm at best for this sort of analysis. The point 
here echoes a point like the one made in the Spring 19 66 lectures apropos of 
“possessive pronouns;” they work linguistically as possessives only given an 
independent analysis of what they are affiliated to as “possessable” (hence the 
very different senses of “my shoes” and “my barber”). And the latter are not 
linguistic facts. 30 

But more is involved than there simply being a separate domain to be 
studied here, and therefore possibly a different discipline. When he turns to 
Bernice’s answer, Sacks notes that it says “The question you asked me is 
correct. I am what you’re supposing I am.” Then (p. 539): 

And by using “by far” one indicates how one would know it; i.e., by 
looking around the class, without any particular interest in finding out 
the ages, she could age herself relatively to everyone else - which is after 
all not a thing that many in a class would do. But there are some people 
who can do it just like that, by virtue of that it’s a ‘by far.’ That is to 
say, ‘by far’ is glance-determinable . And if it’s glance-determinable, then 
that’s how you could have known it . . . It’s visible, like anything else 

30 This discussion is clearly related closely to the one about “frame-and-slot” analysis in 
lecture 1 of the Winter 1970 set, and cf. above at pp. xxi-xxii. 



in the room, that she is older by far. And as she knows it, so does 

anybody else in the class know it. 

Sacks then points out that “that the answer says how one knows what one is 
saying is a common feature of answers.” This is the sort of thing that linguists 
(e.g., Chafe and Nichols, 1986) mean by the term “evidentials.” But “by 
far” is not, I believe, the sort of item (such as modals like “must have,” 
attributions like “John said . . . ,” access routes such as “I read that ...” etc.) 
that is ordinarily counted as an evidential. It is not a linguistic feature, but a 
grasp of the course of action by which such a formulation would come to be 
made, and via an appreciation of its consequentiality to the circumstances of 
the one making it, that “by far the oldest” as glance-determinable needs to 
be understood. For while “by far” may have these attributes for this question 
by this asker to this recipient about this setting, it is by no means a feature of 
its linguistic realization per se, or even one of its variants. The range of further 
observations which an exchange like this can engender, and the theoretical 
directions in which they lead (both of which Sacks pursues in the remainder 
of this discussion) belong to a domain of inquiry that may well be a necessary 
complement to a thoroughgoing linguistics but is not part of it, and should 
be part of a thoroughgoing sociology or anthropology, but does not seem 
likely to become that either. 

The Spring 1972 lectures present various of the juxtapositions or contrasts 
which run through Sacks’ oeuvre. Lectures 1 and 2 juxtapose discussions of the 
most abstract and general sort - characterizing a formal structure, the 
adjacency pair, not only as a type or class of occurence, but as a class of classes 
- with a detailed examination of a single small excerpt from a conversation 
which is turned into a window through which the phenomenology (in a 
non-technical sense) of a person’s social circumstances and experience is 
captured and fleshed out in a compelling fashion, and in a manner which 
resonates to the circumstances and experiences of many who might find 
themselves in cognate circumstances. 

Lecture 3 begins with another excerpt, but uses it largely as the point of 
departure for a discussion of a type of sequence and of a characterizable locus 
of interactional experience - the initial contact between someone calling on the 
telephone and someone answering. The launching of the discussion from a 
particular exemplar of an opening sequence imparts the flavor of empirical 
analysis to the discussion, but in fact it is mainly near the end of the discussion 
that Sacks takes up particulars of that initial fragment. In between his 
characterization is chock full of the products of many empirical analyses, but 
only their upshots are offered, with intendedly typicalized reports of conver- 
sational exchanges to instantiate the themes, rather than analysis in each case, 
for each observation or upshot, with specific instances or exemplars. Here 
again Sacks catches the phenomenology of a social-interactional place in the 
world, but whereas the place in the first two lectures was something like 
“being a certain type of unique person in a setting,” here it is a transient 
(though potentially recurrent) interaction state - answerer of the phone who 

Introduction xlvii 

may or may not be the “called,” and, if not, who may or may not get talked 
to. 31 

Lecture 4 is a specially striking exemplification of Sacks’ ability to 
formulate an absolutely abstract issue, problem, or way of conceiving the 
organization of talk, and then to use it to set a vernacularly characterizable 
and recognizable class of occurrences into a relevant theoretical “space.” Here, 
Sacks proposes to reconceive all utterances in (a) conversation in terms of three 
possible “positions:” last, current, next, and he then begins a course of 
theoretical observations about one of them - “next position” - as a purely 
abstract possible object; and he finds, given how conversation seems empiri- 
cally to be organized (especially given the turn-taking system which it seems 
to employ) a set of characteristics of “next position” per se, characteristics 
which will always have some particular embodiment by virtue of the 
particular “current” utterance relative to which another is “next,” but which 
are features of “next” position generically. From one such set of features - 
that any “next” can accommodate some range of possible utterances or 
utterance types, but not any utterance or utterance types - Sacks shows how 
competition for a turn falls out as a consequence. For a possible next speaker 
with something particular to say may see that it is possible to say it “next,” 
but that each future “current” may restrict against this sayable in its next 
position. Were things otherwise organized, a speaker with something to say 
would never need to get a particular next position to say it in at the cost of 
not getting to say it; everything “intended to be said” could, and perhaps 
would, get said eventually — in some “any” next turn. 

The power of this analytic tool is potentially very extensive, and some of it 
made its way into the eventually published version of the turn-taking paper 
(Sacks, Schegloff and Jefferson, 1974). More work along these lines was 
planned; perhaps some day more will appear, however impoverished by 
Sacks’ unavailability to press it ahead in his own distinctive way. 


After the Spring term of 1972 Sacks no longer recorded his lectures, and 
made no special provision for circulating the work which he was teaching in 

3 'i should remark that in this lecture - lecture 3 - more than any other place in the lectures, 
there is a dialogue going on between Sacks and myself - my own part in it having been 
developed first in my dissertation (1967) and the initial paper (Schegloff, 1968) drawn from 
it, and then, most proximately to this lecture, in a revision of several chapters of the 
dissertation for possible book publication, undertaken in the summer of 1970, and discussed 
extensively with Sacks at the end of that summer. Some of that work has subsequently 
appeared in modified form, e.g., in Schegloff, 1979 and 1986. 

32 As noted earlier, at least some of Sacks’ lectures at the Linguistic Institute held during the 
summer of 1973 were recorded, though Sacks did not choose to have them transcribed for 
circulation. As well, Sacks continued to record many seminars and working sessions with 
students and colleagues. 



his classes. As it happened, I was that summer moving from a position at 
Columbia University to one at UCLA, and for the next three years Sacks and 
I maintained an often intensive, and intermittently attenuated, period of 
collaborative work. Most of both Sacks’ sole-authored work and mine which 
appeared over the following half dozen years was the delayed publication of 
work done and written up much earlier. 33 

Leaving aside for the moment work that was being newly launched or 
developed in fresh directions during the years from Fall 1972 to Sacks’ death 
in November, 1975, 34 those years saw the drafting of the paper on 
turn-taking (Sacks, Schegloff and Jefferson, 1974), the earlier-mentioned 
‘Two preferences . . .’ paper (Sacks and Schegloff, 1979), a paper on laughter 
(Jefferson, Sacks and Schegloff, 1984) and the paper on repair (Schegloff, 
Jefferson and Sacks, 1977). 35 An extensive account of the foci of work during 
these years is beyond the scope of the present introduction. 36 


During the winter of 1974-5, Sacks and I were approached by several faculty 
members at the University of California, Santa Barbara about the possibility 
of establishing there an interdisciplinary program focussed on language, 
discourse and interaction. We explored the possibility through the first half of 
1975; we each visited the campus, gave talks, discussed the prospects with 
the local interested faculty. It seemed increasingly clear that this was a serious 
possibility, and that what was wanted was just the sort of enterprise that 
conversation analysis was becoming - had already become. The prospects 

33 Thus: Sacks, 1972a was the published version of Sacks' end product, dated June, 1965, 
of what (rendered in more accessible language by David Sudnow) was Sacks’ Ph.D. 
dissertation. Sacks, 1972b was a somewhat edited version of lectures from 19 66. Sacks, 
1972c was originally a graduate student paper, written in 1962-3. Sacks, 1973 was the 
published version of Sacks’ paper at the Georgetown Roundtable held in March, 1972. Sacks, 
1974 was the published version of a paper delivered at a conference held in April, 1972. 
Sacks, 1975 was the edited version of a lecture last given in 1968. Subsequent publications 
under his name are edited versions of all or parts of pre-1972 course lectures, assembled by 
Gail Jefferson (cf. introduction to Volume 1, p. ix, n. 1). Only Sacks, 1987 [1973], although 
edited by others from a lecture, was first delivered after spring 1972. 

Of co-authored papers, Schegloff and Sacks, 1973 was drafted (substantially in the form 
in which it was published) in 1969. 

34 Including his beginning to work with video materials, prompted in part during the 1973 
Linguistic Institute by seeing the work of Charles and Marjorie Goodwin and its fit to 
conversation analytic concerns. 

35 Which Sacks and I outlined together in the spring of 1975, and which I then wrote the 
intial draft of, after Sacks went off to the first Boston University Summer Session on 
Ethnomethodology and Conversation Analysis. 

36 After his death, I made a list of papers we had discussed more than once, and more than 
casually, as needing to be drafted. There were 26 of them. Some account of these years may 
yet be written. 



became increasingly enhanced. Jefferson (on the UCSB faculty that year) 
could also be appointed; we could tailor a curriculum to the special character 
of the subject matter and our approach to it; scheduling could be made 
flexible; space for a collegial, “working group” arrangement was possible; 
there would be support for our equipment needs, etc. 

Finally, in the Fall of 1975, we received a request from those who were 
guiding these developments at Santa Barbara. They wanted - quite reason- 
ably - to know what we proposed to offer as a program in return for the 
resources and possibilities which had been discussed over the preceding 
months. Sacks and I had several informal conversations about this. Finally, in 
mid-November, we decided we really had to sit down together and draw up 
a serious plan to offer to Santa Barbara. We decided to meet at Irvine, in part 
because Sacks had been suffering from an ear infection. We tentatively agreed 
to meet on a Monday morning. When I called the Sunday night before the 
scheduled meeting, the infection had not yet fully cleared up, and Sacks was 
still taking medication for it. But he resisted the suggestion that we delay the 
meeting. We would meet at the Irvine campus. 

It was on his drive from the back canyons of Orange County to the Irvine 
campus to discuss the specifics of the program in conversation analysis which 
we might propose to Santa Barbara that his car was involved in a head-on 
collision with a truck, and he was killed. 


Reading the lectures now, and especially reading ones which entertain agendas 
of work to be done (e.g., the last pages of Spring 1972, lecture 5), poses again 
and again the question of where our understanding of language and talk, of 
interaction and the social fates played out in it, of human sociality from the 
most intimate emotion to the largest issues of social organization, where our 
understanding would now be had Sacks not died in November, 1975. 
Recalling the years immediately following the last of these lectures, when 
some of that work was being advanced, and imagining what might have been 
accomplished in a program designed to advance this undertaking, in a 
supportive institutional environment, enhances the fantasy. 

Whether or not the efforts of others succeed in establishing a discipline with 
satisfactory payoffs and sustainable continuity, we shall not have the 
discipline, or the understanding, which we would have had with him. Nor 
will it avail for others literally to try to execute the plans of inquiry which he 
projected. They were built from the breadth of his own past reading, from the 
depth and range of his analytic and empirical work, and were the product of 
the very special metier of his mind. What is needed is a continuous re- 
energizing of inquiry by the example of his work and the possibilities which it 
revealed - each person bringing to the enterprise the best mastery of past work 
which they can achieve and the special contribution which the character of their 
own talent makes possible. Not mechanical imitation or extrapolation but the 



best possible effort to advance the undertaking in original ways will constitute 
the most appropriate and enduring celebration of Sacks’ contribution. 

The first lecture presented in these volumes began with a consideration of 
a conversation’s opening; the last ends with a puzzle about how much can be 
infused into a conversation-opening “hello.” An astonishingly rich tapestry of 
analysis comes between, in an intellectual career which did not tire of 
repeatedly going back to the beginning, showing again and again that there 
was an enterprise to be undertaken here. The achievement of the work is to 
be found not only in its results, but in its prompting of an undertaking, and 
in its constituting a standing invitation to others to join, and to begin, that 
undertaking themselves. 




Albert, Ethel, 1964. Rhetoric, logic, and poetics in Burundi: cultural patterning of speech 
behavior. American Anthropologist, 66, 6, part 2. 

Atkinson, J. Maxwell and Drew, Paul, 1979. Order in Court (London: Macmillan). 

Atkinson, J. Maxwell and Heritage, John C. (eds), 1984. Structures of Social Action 
(Cambridge: Cambridge University Press). 

Barker, Roger G. and Wright, HerbertF., 1951. One Boy's Day (New York: Harper and 

Bilmes, Tack, 1988. The concept of preference in conversation analysis. Language in 
Society, 17, 161-81. 

Chafe, Wallace and Nichols, Johanna (eds), 1986. Evidentiality: the linguistic coding of 
epistemology (Norwood, NJ: Ablex Publishing). 

Davidson, Judy, 1984. Subsequent versions of invitations, offers, requests and proposals 
dealing with potential or actual rejection. In JM Atkinson and JC Heritage (eds), 
Structures of Social Action (Cambridge: Cambridge University Press). 

Davidson, Judy, 1990. Modifications of invitations, offers and rejections. In George 
Psathas (ed.), Interaction Competence (Washington, DC: University Press of America). 

Garfinkel, Harold, 1967. Studies in Ethnomethodology ( Englewood Cliffs, NJ: Prentice 

Gellner, Ernest, 1975. Ethnomethodology: the re-enchantment industry or a Californian 
way of subjectivity. Philosophy of the Social Sciences, 5, 4, 431-50. 

Goffman, Erving, 1983. The interaction order. American Sociological Review, 48, 1-17. 

Havelock, Eric A., 1963. Preface to Plato (Cambridge, Mass.: Harvard University Press). 

Heritage, John C., 1984. Garfinkel and Ethnomethodology (Cambridge: Polity Press). 

Homans, George C., 1961. Social Behavior: its elementary forms (New York: Harcourt, 
Brace and World). 

Jefferson, Gail, Sacks, Harvey, and Schegloff, Emanuel A., 1984. On laughter in pursuit of 
intimacy. Working Papers, C(135), 1-20. Centro Intemazionale de Semistica e di 
Linguistica, Urbino, Italy, Full version in Graham Button and John R. E. Lee (eds.) Talk 
and Social Organization. 1987 (Clevedon, UK: Multi-liugual Matters) pp. 152-205. 

Kroeber, Alfred, 1917. The superorganic. American Anthropologist, 19, 163-213. 

Levinson, Stephen C., 1983. Pragmatics (Cambridge: Cambridge University Press). 

Lord, Albert Bates, I960. The Singer of Tales (Cambridge, Mass: Harvard University 

McLellan, David, 1977. Karl Marx: selected writings (Oxford: Oxford University Press). 

Parry, Milman, 1971. The Making of Homeric Verse: the collected papers of Milman Parry, 
edited by Adam Parry. (Oxford: The Clarendon Press). 

Pomerantz, Anita, 1978. Compliment responses: notes on the co-operation of multiple 
constraints. In Jim Schenkein (ed.). Studies in the Organization of Conversational 
Interaction (New York: Academic Press), 79-112. 

Pomerantz, Anita, 1984. Agreeing and disagreeing with assessments: some features of 
preferred /dispreferred turn shapes. In J. M. Atkinson and J. C. Heritage (eds), 
Structures of Social Action (Cambridge: Cambridge University Press), 57-101. 

Sacks, Harvey, 1963. Sociological description. Berkeley Journal of Sociology, 8, 1-16. 

Sacks, Harvey, 1972a. An initial investigation of the usability of conversational data for 
doing sociology. In David N. Sudnow (ed.), Studies in Social Interaction (New York: 
The Free Press), 31-74. 

Sacks, Harvey, 1972b. On the analyzability of stories by children. In John J. Gumperz 
and Dell Hymes (eds). Directions in Sociolinguistics (New York: Holt, Rinehart and 
Winston), 325-45. 



Sacks, Harvey, 1972c. Notes on police assessment of moral character. In David N. 
Sudnow (ed.), Studies in Social Interaction (New York: The Free Press), 280-93. 

Sacks, Harvey, 1973. On some puns with some intimations. In Roger W. Shuy (ed.). 
Report of the 23rd Annual Roundtable Meeting on Linguistics and Language Studies 
(Washington, DC: Georgetown University Press), 135-44. 

Sacks, Harvey, 1974. An analysis of the course of a joke’s telling in conversation. In 
Richard Bauman and Joel Sherzer (eds), Explorations in the Ethnography of Speaking 
(Cambridge: Cambridge University Press), 337-53. 

Sacks, Harvey, 1975. Everyone has to lie. In Ben Blount and Mary Sanchez (eds), 
Sociocultural Dimensions of Language Use (New York: Academic Press), 57-80. 

Sacks, Harvey, 1978. Some technical considerations of a dirty joke. In Jim Schenkein 
(ed.), Studies in the Organization of Conversational Interaction (New York: Academic 
Press), 249-70 [edited by Gail Jefferson from unpublished lectures]. 

Sacks, Harvey, 1987 [1973]. On the preferences for agreement and contiguity in 
sequences in conversation. In Graham Button and John R. E. Lee (eds), 1987. Talk 
and Social Organization (Clevedon: Multilingual Matters), 54-69. 

Sacks, Harvey and Schegloff, Emanuel A., 1979. Two preferences in the organization of 
reference to persons in conversation and their interaction. In George Psathas (ed.), 
Everyday Language: studies in ethnomethodology (New York: Irvington), 15-21. 

Sacks, Harvey, Schegloff, Emanuel A., and Jefferson, Gail, 1974. A simplest systematics 
for the organization of turn-taking for conversation. Language, 50, 696-735. A 
variant version also published in Jim Schenkein (ed.), Studies in the Organization of 
Conversational Interaction (New York: Academic Press, 1978). 

Schegloff, Emanuel A., 1967. The First Five Seconds: the order of conversational openings. 
PhD dissertation, Department of Sociology, University of California, Berkeley. 

Schegloff, Emanuel A., 1968. Sequencing in conversational openings. American Anthro- 
pologist, 70, 1075-95. 

Schegloff, Emanuel A., 1970. Openings sequencing and Answers. In The Social 
Organization of Conversational Openings (MS). 

Schegloff, Emanuel A., 1979. Identification and recognition in telephone conversation 
openings. In George Psathas (ed.), Everyday Language: studies in ethnomethodology 
(New York: Irvington). 

Schegloff, Emanuel A., 1986. The routine as achievement. Human Studies, 9, 111-51. 

Schegloff, Emanuel A., 1988. On an actual virtual servo-mechanism for guessing bad 
news: a single case conjecture. Social Problems, 35, 442-57. 

Schegloff, Emanuel A. and Sacks, Harvey, 1973. Opening up closings. Semiotica, 8, 

Schegloff, Emanuel A., Jefferson, Gail, and Sacks, Harvey, 1977 . The preference for 
self-correction in the organization of repair in conversation. Language, 53, 361-82. 

Schenkein, Jim (ed.), 1978. Studies in the Organization of Conversational Interaction 
(New York: Academic Press). 

Searle, John, 1986. Introductory essay: notes on conversation. In Donald G. Ellis and 
William A Donahue (eds), Contemporary Issues in Language and Discourse Processes 
(Hillsdale, NJ: Lawrence Erlbaum Associates). 

Searle, John, 1992. Conversation reconsidered. In John R. Searle et al., (On) Searle on 
Conversation (Amsterdam /Philadelphia: John Benjamin). 

Taylor, Archer, 1931. The Proverb (Cambridge, Mass: Harvard University Press). 

A Note on Some Effects of the 
Taping, Transcribing, Editing, and 
Publishing upon the Materials 

Each process that the lectures have gone through, from taperecording to 
typesetting, has had its effect on what appears in these volumes. As the 
lectures were being taped, faulty equipment, excessive background noise, etc., 
could result in gaps within a recording, or the disappearance of an entire 
lecture. This has provided an opportunity for the presentation of samples of 
another instrument, as some of the absences have been filled in with materials 
from Sacks’ research notes. 

The transcribing process generated shoals of mis-hearings, only some of 
which have been corrected. For example, a rendering of ‘Oral styles of 
American folk narrators’ as “All styles . . . ”, or the transformation somehow 
of ‘Von Neumann’ to “Baganin” were caught as the bibliographical 
references were being assembled. A description, “the police car . . . follows 
him along for a mile, finally pulls him over,” emerged on some nth reading 
as much more likely to be “ . . . follows him along for awhile.” And a fresh 
and knowledgable eye discovered that “. . . and they had a fantastic scene 
with persons coming in ...” was almost certainly “. . . and they had a 
fantastic team of persons coming in,” and rather more consequentially, that 
“That’s in part the problem with Reichenbach’s second chapter” is in fact 
reference to his “seventh” chapter, that a murky reference to data consisting 
of “whatever it is that we have to have” was surely talking about “whatever 
it is that we happen to have”, and reference to “the attending of a 
prepositional phrase” was obviously reference to the “appending” of the 
phrase. God only knows what further errors have slipped through. 

There was never any requirement that the transcripts be verbatim, and 
there is variation across and within them, although one small batch was 
produced with a systematic concern for the very words. Not long after Sacks 
was killed, a cache of taperecordings of his earliest lectures turned up. These 
were treated, not as usual working tapes (to be transcribed as quickly as 
possible and then tossed back into the pool of tapes for reuse), but as 
something approaching a memorial. At attempt was made to capture as much 
detail as possible; i.e., to transcribe them at the level of detail used on the 
research materials, with Sacks’ frequent and prolonged silences, long drawn 
out “uh”s, and very slight New York accent faithfully notated. But the 
attempt was abandoned in the middle of the second page of transcript: At 
that level of detail the lecture was simply not followable. It was necessary to 
return to the standard format of the workaday lecture transcripts - the sense 




of the specialness of this particular batch of tapes relegated to a commitment 
to word for word accuracy. For the most part, however, some degree ol 
spontaneous editing occurred at the transcription stage. 

The editing per se ranges from faithful conservation of the text to 
something very like wanton tampering. Now and again, remarks on 
particular editorial effects can be found (see pp. ix-xi, 126-3 1, and 507-12), 
but there is no comprehensive discussion. 

As the volumes went to press, another series of changes occurred, geared to 
bringing the materials more into line with standard literary usage, for 
example, replacement of devices intended to emphasize the fact that these are 
transcripts of spontaneous talk rather than written text; e.g., the rendering of 
‘etc.’ as “etcetera” and (rather more variably) the rendering of numbers as 
words, by the standard abbreviations and numerals. And, for example, the 
various references to books and articles now look very much more like 
standard bibliographical notes than spoken citations. A reference which went 
into the typescript as “I come by this sort of consideration via a paper written 
by a fellow named Richard Gunter, 19 66, Journal of Linguistics called ‘On 
accents in dialogue,’” became “I come by this sort of consideration via a 
paper by Richard Gunter, ‘On accents in dialogue,’ Journal of Linguistics 
(1966).” A typescript entry, reference to a review article “in a book called 
Studies in Language and Literature, edited by A. Marquart, 1954, called 
‘English sentence connectors,’ by Seymour Chapman. That’s on page 315,” 
came back as reference to a review article “by Seymour Chapman, ‘English 
sentence connectors,’ in A. Marquart (ed.), Studies in Language and Literature 
(1954), p. 315.” As a sort of compromise between good form and actual 
occurrence, the non-initial reference to the author, in this and other citations 
throughout the volumes, has been restored. 


With thanks to David Sudnow who kick-started the editing process when it 
had stalled, and to Betty Sacks who kept it going. And to lan Anderson, J. 
Maxwell Atkinson, Tom Boda, Judy Davidson, Paul Drew, Robert Dunstan, 
Konrad Ehlich, B.J. Fehr, Rich Frankel, Jo Ann Goldberg, Charles Goodwin, 
Marjorie Goodwin, Auli Hakulinen, Paul ten Have, Christian Heath, John 
Heritage, Hanneke Houtkoop, Willem Houtkoop, Eva Konig, Jerry 
Krakowski, John R.E. Lee, Diane Lee, Gene Lerner, Jennifer Mandelbaum, 
Michael Moerman, Bruce Peddy, Anita Pomerantz, George Psathas, Blaine 
Roberts, Joan Sacks, Jim Schenkein, Albert Stuulen, Alene Terasaki, Roy 
Turner, Rod Watson, and those known only as ‘transcriber unknown’. 

The editor and publishers wish to thank the following for permission to use 
copyright material: Macmillian Publishing Company for an extract from The 
Urban Villagers by Herbert J. Gans; Mouton de Gruyter for an extract from 
Language in the Crib by Ruth Weir; Oxford University Press for an extract 
from ‘Is everyday language inconsistent?’ by Avram Stroll, in Mind, LXIII, 
no. 250. Every effort has been made to trace copyright holders of extracted 
material and the publishers would be grateful to hear from any individuals or 
institutions whose clearance has not been obtained at the date of publication. 


Part I 

Fall 1964-Spring 1965 

Lectures on Conversation, Volume I, II Harvey Sacks 
© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6 

Lecture 1 

Rules of Conversational Sequence 

I’ll start off by giving some quotations. 

(1) A: Hello 
B: Hello 

(2) A: This is Mr Smith may I help you 
B: Yes, this is Mr Brown 

(3) A: This is Mr Smith may I help you 
B: I can’t hear you. 

A: This is Mr Smith . 

B: Smith. 

These are some first exchanges in telephone conversations collected at an 
emergency psychiatric hospital. They are occurring between persons who 
haven’t talked to each other before. One of them, A, is a staff member of this 
psychiatric hospital. B can be either somebody calling about themselves, that 
is to say in trouble in one way or another, or somebody calling about 
somebody else. 

I have a large collection of these conversations, and I got started looking at 
these first exchanges as follows. A series of persons who called this place 
would not give their names. The hospital’s concern was, can anything be done 
about it? One question I wanted to address was, where in the course of the 
conversation could you tell that somebody would not give their name? So I 
began to look at the materials. It was in fact on the basis of that question that 
I began to try to deal in detail with conversations. 

I found something that struck me as fairly interesting quite early. And that 
was that if the staff member used “This is Mr Smith may I help you” as their 
opening line, then overwhelmingly, any answer other than “Yes, this is Mr 
Brown” (for example, “I can’t hear you,” “I don’t know,” “How do you 

A combination of Fall 1964, tape 1, side 2 and tape 2, side 1, with brief extracts 
from Winter 1965, lecture (1) - the parenthesis indicate that the original transcripts 
were unnumbered, the current numbering likely but not certain - pp. 1 and 11-12 
(transcriber unknown) and Spring 1965 (’64-’65), lecture 3, pp. 6-7 (transcriber 

The lectures’ titles are intended to give a handle on them, and only partially 
capture the contents. 

Lectures on Conversation, Volume I, II Harvey Sacks 
© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6 



Part I 

spell your name?”) meant that you would have serious trouble getting the 
caller’s name, if you got the name at all. 

I’m going to show some of the ways that I’ve been developing of analyzing 
stuff like this. There will be series of ways fitted to each other, as though one 
were constructing a multi-dimensional jigsaw puzzle. One or another piece 
can be isolated and studied, and also the various pieces can be studied as to 
how they fit together. I’ll be focussing on a variety of things, starting off with 
what I’ll call ‘rules of conversational sequence.’ 

Looking at the first exchange compared to the second, we can be struck by 
two things. First of all, there seems to be a fit between what the first person 
who speaks uses as their greeting, and what the person who is given that 
greeting returns. So that if A says “Hello,” then B tends to say “Hello.” If 
A says“This is Mr Smith may I help you,” B tends to say “Yes, this is Mr 
Brown.” We can say there’s a procedural rule there, that a person who speaks 
first in a telephone conversation can choose their form of address, and in 
choosing their form of address they can thereby choose the form of address the 
other uses. 

By ‘form’ I mean in part that the exchanges occur as ‘units.’ That is, 
“Hello” “Hello” is a unit, and “This is Mr Smith may I help you” “Yes, this 
is Mr Brown” is a unit. They come in pairs. Saying “This is Mr Smith may 
I help you” thereby provides a ‘slot’ to the other wherein they properly would 
answer “Yes, this is Mr Brown.” The procedural rule would describe the 
occurrences in the first two exchanges. It won’t describe the third exchange, 
but we’ll come to see what is involved in such materials. 

Secondly, if it is so that there is a rule that the person who goes first can 
choose their form of address and thereby choose the other’s, then for the unit, 
“This is Mr Smith may I help you” “Yes, this is Mr Brown,” if a person uses 
“This is Mr Smith ...” they have a way of asking for the other’s name - 
without, however, asking the question, “What is your name?” And there is 
a difference between saying “This is Mr Smith may I help you” - thereby 
providing a slot to the other wherein they properly would answer “Yes, this 
is Mr Brown” - and asking the question “What is your name?” at some point 
in the conversation. They are very different phenomena. 

For one, in almost all of the cases where the person doesn’t give their name 
originally, then at some point in the conversation they’re asked for their name. 
One way of asking is just the question “Would you give me your name?” To 
that, there are alternative returns, including “No” and “Why?” If a caller 
says “Why?” the staff member may say something like, “I want to have 
something to call you” or “It’s just for our records.” If a caller says “No,” 
then the staff member says “Why?” and may get something like “I’m not 
ready to do that” or “I’m ashamed.” 

Now, I’ll consider many times the use of “Why?” What I want to say 
about it just to begin with, is that what one does with “Why?” is to propose 
about some action that it is an ‘accountable action.’ That is to say, “Why?” 
is a way of asking for an account. Accounts are most extraordinary. And the 
use of accounts and the use of requests for accounts are very highly regulated 

Lecture 1 


phenomena. We can begin to cut into these regularities by looking at what 
happens when “May I have your name?” is followed by “Why?” Then you 
get an account; for example, “I need something to call you.” The other might 
then say, “I don’t mind.” Or you might get an account, “It’s just for our 
records.” To which the other might say, “Well I’m not sure I want to do 
anything with you, I just want to find out what you do” - so that the records 
are not relevant. 

What we can see is that there are ways that accounts seem to be dealable 
with. If a person offers an account, which they take it provides for the action 
in question being done - for example, the caller’s name being given - then if 
the other can show that the interest of that account can be satisfied without 
the name being given, the name doesn’t have to be given. That is, if the 
account is to control the action, then if you can find a way that the account 
controls the alternative action than it proposed to control, you can use it that 

It seems to be quite important, then, who it is that offers the account. 
Because the task of the person who is offered the account can then be to, in 
some way, counter it. Where, alternatively, persons who offer an account seem 
to feel that they’re somehow committed to it, and if it turns out to be, for 
example, inadequate, then they have to stand by it. 

The fact that you could use questions — like “Why?” — to generate 
accounts, and then use accounts to control activities, can be marked down as, 
I think, one of the greatest discoveries in Western civilization. It may well be 
that that is what Socrates discovered. With his dialectic he found a set of 
procedures by which this thing, which was not used systematically, could 
become a systematic device. Socrates will constantly ask “Why?,” there will 
be an answer, and he’ll go on to show that that can’t be the answer. And that 
persons were terribly pained to go through this whole business is clear enough 
from the Dialogues. And it’s also clear in our own experiences. And in the 
materials I’ll present. 

We see, then, one clear difference between providing a slot for a name, and 
asking for a name. Asking for a name tends to generate accounts and 
counters. By providing a slot for a name, those activities do not arise. 

We can also notice that, as a way of asking for the other’s name, “This is 
Mr Smith . . .” is, in the first place, not an accountable action. By that I mean 
to say, it’s not required that staff members use it and they don’t always use 
it, but when they do, the caller doesn’t ask why. “This is Mr Smith . . .” gets 
its character as a non-accountable action simply by virtue of the fact that this 
is a place where, routinely, two persons speak who haven’t met. In such places 
the person who speaks first can use that object. And we could say about that 
kind of item that the matters discriminated by its proper use are very 
restricted. That is to say, a call is made; the only issue is that two persons are 
speaking who presumably haven’t met, and this object can be used. 

Furthermore, the matters are discriminated in different terms than those 
which the agency is constructed for. That is, they are discriminated in terms 
of ‘two people who haven’t met’ rather than, for example, that an agency staff 


Part 1 

member is speaking to someone calling the agency for help. And where one 
has some organization of activities which sets out to do some task - and in this 
case it’s important for the agency to get names - then if you find a device 
which discriminates in such a restricted fashion, you can use that device to do 
tasks for you. 

Now, given the fact that such a greeting as “This is Mr Smith 
provides for the other giving his own name as an answer, one can see what the 
advantage of “Hello’’ is for someone who doesn’t want to give their name. 
And I found in the first instance that while sometimes the staff members use 
“Hello’’ as their opening line, if it ever occurred that the persons calling the 
agency spoke first, they always said “Hello.” 

Persons calling could come to speak first because at this agency, caller and 
staff member are connected by an operator. The operator says “Go ahead 
please” and now the two parties are on an open line, and one can start talking 
or the other can start talking. This stands in contrast to, for example, calling 
someone’s home. There, the rights are clearly assigned; the person who 
answers the phone speaks first. If they speak first, they have the right to 
choose their form. If they have the right to choose their form, they have the 
right thereby to choose the other’s. Here, where the rights are not clearly 
assigned, the caller could move to speak first and thereby choose the form. 
And when callers to this agency speak first, the form they choose is the unit 
“Hello” “Hello.” Since such a unit involves no exchange of names, they can 
speak without giving their name and be going about things in a perfectly 
appropriate way. 

Now, there are variant returns to “This is Mr Smith may I help you?” one 
of which is in our set of three exchanges: “I can’t hear you.” I want to talk 
of that as an ‘occasionally usable’ device. That is to say, there doesn’t have to 
be a particular sort of thing preceding it; it can come at any place in a 
conversation. Here is one from the middle of a conversation, from a different 
bunch of materials. 

A: Hey you got a cigarette Axum. I ain’t got, I ain’t got a good cigarette, 

and I can’t roll one right now. Think you can afford it maybe? 

B: I am not here to support your habits. 

A: Huh? My helplessness? 

B: I am not responsible for supporting your habits ( ) 

A: My habits ((laughing)) 

Our third exchange from the psychiatric hospital has the device used at the 
beginning of the conversation. 

A: This is Mr Smith may I help you 

B: I can’t hear you. 

A: This is Mr Smith . 

B: Smith. 

Lecture 1 


What kind of a device is it? What you can see is this. When you say “I 
can’t hear you,” you provide that the other person can repeat what they said. 
Now what does that repetition do for you? Imagine you’re in a game. One of 
the questions relevant to the game would be, is there a way in that game of 
skipping a move? It seems that something like ”1 can’t hear you” can do such 
a job. If you introduce it you provide for the other to do some version of a 
repeat following which you yourself can repeat. And then it’s the other’s turn 
to talk again. What we find is that the slot where the return would go - your 
name in return to “This is Mr Smith ...” - never occurs. 

It is not simply that the caller ignores what they properly ought to do, but 
something rather more exquisite. That is, they have ways of providing that 
the place where the return name fits is never opened. So that their name is not 
absent. Their name would be absent if they just went ahead and talked. But 
that very rarely occurs. The rules of etiquette - if you want to call them that, 
though we take etiquette to be something very light and uninteresting and to 
be breached as you please - seem to be quite strong. Persons will use ways to 
not ignore what they properly ought to do by providing that the place for 
them to do it is never opened. 

I hope it can also be seen that a device like “I can’t hear you” - the repeat 
device, providing for a repetition of the thing that was first said, which is then 
repeated by the person who said “I can’t hear you” - is not necessarily 
designed for skipping a move. It is not specific to providing a way of keeping 
in the conversation and behaving properly while not giving one’s name. It can 
be used for other purposes and do other tasks, and it can be used with other 
items. That’s why I talk about it as an ‘occasional device.’ But where that is 
what one is trying to do, it’s a rather neat device. 

Let me turn now to a consideration which deals with a variant return to 
“May I help you?” That is, not “Yes . . .” but “I don’t know.” I’ll show a 
rather elaborate exchange in which the staff member opens with a version of 
“This is Mr Smith may I help you” but the combination gets split. The name 
is dealt with, and when the “can I help you” is offered, it occurs in such a way 
that it can be answered independent of the name. 1 

Op : Go ahead please 

A: This is Mr Smith (13: Hello) of the Emergency Psychiatric Center can 

I help you. 

B: Hello? 

A: Hello 

B: I can’t hear you. 

A: I see. Can you hear me now? 

B: Barely. Where are you, in the womb? 

1 The fragment of data is reproduced pretty much as Sacks transcribed it, 
preserving his attempts to deal with simultaneous talk (i.e., A: This is Mr Smith ( B : 
Hello) of the Emergency Psychiatric Center) and silence (e.g., B: I uh Now that 
you’re here . . .). See lecture 9, pp. 66 and 68 for two other approaches by him to 
simultaneous talk in this same conversation. 


Part I 

A: Where are you calling from? 

B: Hollywood. 

A: Hollywood. 

B: I can hear you a little better. 

A: Okay. Uh I was saying my name is Smith and I’m with the 

Emergency Psychiatric Center. 

B : Your name is what? 

A: Smith. 

B : Smith? 

A: Yes. 

A: Can I help you? 

B\ I don’t know hhheh I hope you can. 

A: Uh hah Tell me about your problems. 

B: I uh Now that you’re here I’m embarassed to talk about it. I don’t 

want you telling me I’m emotionally immature ’cause I know I am 

I was very puzzled by “I don’t know” in return to “May I help you.” I 
couldn’t figure out what they were doing with it. And the reason I was 
puzzled was that having listened to so many of these things and having been 
through the scene so many times, I heard “May I help you” as something like 
an idiom. I’m going to call these idiom-like things ‘composites.’ That means 
you hear the whole thing as a form, a single unit. And as a single unit, it has 
a proper return. As a composite, “May I help you” is a piece of etiquette; a 
way of introducing oneself as someone who is in the business of helping 
somebody, the answer to which is “Yes” and then some statement of what it 
is one wants. We can consider this item in terms of what I’ll call the ‘base 
environment’ of its use. 

By ‘base environment’ I mean, if you go into a department store, 
somebody is liable to come up to you and say “May I help you.” And in 
business-type phone calls this item is routinely used. And if you come into a 
place and you don’t know what it’s like, and somebody comes up to you and 
uses such an item, that’s one way of informing you what kind of a place it is. 
So, if a new institution is being set up, then there are available in the society 
whole sets of ways that persons go about beginning conversations, and one 
could, for example, adopt one or another of a series of them as the ones that 
are going to be used in this place. 

Now the thing about at least some composites is that they can be heard not 
only as composites, but as ordinary sentences, which we could call ‘construc- 
tives,’ which are understood by taking the pieces and adding them up in some 
way. As a composite, “May I help you” is a piece of etiquette, a signal for 
stating your request - what you want to be helped with. Alternatively, as a 
constructive, “May I help you” is a question. If one hears it as a question, the 
piece of etiquette and its work hasn’t come up, and “I don’t know” is a 
perfectly proper answer. 

Further, “I don’t know” may be locating a problem which “May I help 
you” is designed, in the first place, to avoid. In its base environment, for 

Lecture 1 


example a department store, it’s pretty much the case that for a customer, the 
question of whether some person “can help’’ is a matter of the department 
store having made them the person who does that. That is to say, lots of 
things, like telling you whether you can find lingerie in a certain size, is 
something anybody can do, and as long as the department store says this 
person is going to do it, that’s enough. But we’re dealing with a psychiatric 
hospital. In a department store, being selected to do a job and having 
credentials to do it are essentially the same thing. In a psychiatric hospital and 
lots of other places, however, they are very different things. That is, whether 
somebody can help you if you have a mental disorder, is not solved or is not 
even presumptively solved by the fact that they’ve been selected by somebody 
to do that job. The way it’s solved in this society is by reference to such things 
as having been trained in a particular fashion, having gotten degrees, having 
passed Board examinations, etc. 

Now, in the base environment of the use of “May I help you?’’ there is, as 
I say, no difference essentially between having credentials and being selected. 
If one can formulate the matter in a psychiatric hospital such that those things 
come on as being the same, then one needn’t start off by producing one’s 
credentials at the beginning of the conversation. And in my materials, again 
and again, when “May I help you” is used the person calling says “Yes” and 
begins to state their troubles. 

As a general matter, then, one can begin to look for kinds of objects that 
have a base environment, that, when they get used in that environment 
perform a rather simple task, but that can be used in quite different 
environments to do quite other tasks. So, a matter like ‘credentials’ can be 
handled by this “May I help you” device. There will be lots of other devices 
which have a base environment, which do some other task in some other 

Before moving off of “May I help you” I want to mention one other thing 
about it. If the base environment is something like a department store, then, 
when it’s used in other places - for example, a psychiatric hospital - one of 
the pieces of information it seems to convey is that whatever it is you propose 
to do, you do routinely. To whomsoever that calls. That is, it’s heard as a 
standardized utterance. How is that relevant? It can be relevant in alternative 
ways. First of all, it can be a very reassuring thing to hear. Some persons feel 
that they have troubles, and they don’t know if anybody else has those 
troubles; or, if others do have those troubles, whether anybody knows about 
them. If someone knows about them, then there may be a known solution to 
them. Also and relatedly, a lot of troubles - like mental diseases - are things 
that persons feel very ambivalent about. That is, they’re not sure whether it’s 
some defect of their character, or something else. That, in part, is why they’re 
hesitant to talk about it. And it seems that one of the ways one begins to tell 
people that they can talk, that you know what they have and that you 
routinely deal with such matters, is to use manifestly organizational talk. 

“May I help you,” then, can be a reassuring way to begin. It can 
alternatively be something else. Consider the exchange I just showed, in which 


Part I 

such standardized utterances as “May I help you” and “Tell me about your 
problems” are used. 

A: Can I help you? 

B: I don’t know hhheh I hope you can 

A: Uh hah Tell me about your problems 

B: I uh Now that you’re here I’m embarrassed to talk about it. I 

don’t want you telling me I’m emotionally immature ‘cause I know 
I am 

That is, the use of standardized, manifestly organizational talk can provide for 
the person calling that they’re going to get routine treatment. But ‘routine’, 
for them, may not be such a happy thing. Because, for example, they’ve been 
through it before. But they may have gone through it, as psychiatrists would 
say, part way. For example, they were in analysis for three years and ran out 
of money, or the psychiatrist wouldn’t keep them on, or they didn’t want to 
stay. Part way, they may have come to some point in the analysis where they 
‘knew what was wrong with them.’ That is, they knew the diagnostic term. 
But that diagnostic term may have had a lay affiliate. By that I mean, if a 
psychiatrist says you’re regressed, it’s a technical term. But ‘regressed’ is also 
a lay term, and as a lay term it doesn’t have a great deal of attractiveness. If 
one finds oneself living with a lay understanding of such a term, where the 
term is not a very nice thing to have in its lay sense, then when you hear 
someone using such an item as “May I help you,” you can hear that some 
procedure will be gone through, the upshot of which will be the discovery 
of what you ‘already know’ - the knowing of which doesn’t do you any 

Related to that are such things as, some people seem to feel very much 
disturbed about the fact that their relationship to a psychiatrist or to other 
doctors is monetary. What they want, they say, is a personal solution. Ask 
them what they want, “Well, that you don’t have to pay for it.” When they 
hear “May I help you,” they hear ‘a professional.’ But they feel that the way 
you get cured is by getting an affiliation to somebody which is like the 
affiliations that they failed to get in their lives. That is, they may already have 
come to learn from some other psychiatrist that the failure of love by their 
parents is the cause of their troubles. Then, what they come to see is that they 
need the love of somebody else. And they can’t get that from a therapist. 
Because as soon as they don’t pay, that’s the end of the relationship. 

Now let me just make a few general points. Clearly enough, things like 
“This is Mr Smith,” “May I help you?” and “I can’t hear you” are social 
objects. And if you begin to look at what they do, you can see that they, and 
things like them, provide the makings of activities. You assemble activities by 
using these things. And now when you, or I, or sociologists, watching people 
do things, engage in trying to find out what they do and how they do it, one 
fix which can be used is: Of the enormous range of activities that people do, 
all of them are done with something. Someone says “This is Mr Smith” and 

Lecture 1 


the other supplies his own name. Someone says “May I help you’’ and the 
other states his business. Someone says “Huh?” or “What did you say?” or 
“I can’t hear you,” and then the thing said before gets repeated. What we 
want then to find out is, can we first of all construct the objects that get used 
to make up ranges of activities, and then see how it is those objects do get 

Some of these objects can be used for whole ranges of activities, where for 
different ones a variety of the properties of those objects will get employed. 
And we begin to see alternative properties of those objects. That’s one way we 
can go about beginning to collect the alternative methods that persons use in 
going about doing whatever they have to do. And we can see that these 
methods will be reproducible descriptions in the sense that any scientific 
description might be, such that the natural occurrences that we’re describing 
can yield abstract or general phenomena which need not rely on statistical 
observability for their abstractness or generality. 

There was a very classical argument that it would not be that way; that 
singular events were singular events, given a historian’s sort of argument, that 
they just happen and they get more or less accidentally thrown together. But 
if we could find that there are analytically hard ways of describing these things 
— where, that is, we re talking about objects that can be found elsewhere, that 
get placed, that have ways of being used; that are abstract objects which get 
used on singular occasions and describe singular courses of activity - then 
that’s something which is exceedingly non-trivial to know. 

One final note. When people start to analyze social phenomena, if it looks 
like things occur with the sort of immediacy we find in some of these 
exchanges, then, if you have to make an elaborate analysis of it - that is to say, 
show that they did something as involved as some of the things I have 
proposed - then you figure that they couldn’t have thought that fast. I want 
to suggest that you have to forget that completely. Don’t worry about how 
fast they’re thinking. First of all, don’t worry about whether they’re 
‘thinking.’ Just try to come to terms with how it is that the thing comes off. 
Because you’ll find that they can do these things. Just take any other area of 
natural science and see, for example, how fast molecules do things. And they 
don’t have very good brains. So just let the materials fall as they may. Look 
to see how it is that persons go about producing what they do produce. 

Lecture 2 

On Suicide Threats Getting 
Laughed Off 

Here are some lines that occurred in one of the conversations I collected. This 
is a woman talking. 

A: But about two months ago I was still home on uh one Sunday, oh we 

had five children and I got home from church and he’s got a butcher 
knife. He told the kids to go to the park and play. This is kind of 
unusual for him because he doesn’t like them, especially the baby, to 
go anywhere unless we’re there. 

B- Aha. 

A: After they were all gone, I was laying on the couch just reading the 

Sunday paper and he came over there and started holding this butcher 
knife at my throat. And I said what is the matter with you. He said 
I’m going to kill you. I’m going to end it all. And I said oh for 
goodness sake put it down and go. 

— H started to laugh it ofF. And he sat there for about an hour. So I 
thought well, he kept threatening to kill me. And then he would pull 
it back as if to stab me. And I just laid there and prayed. I almost 
believed he was crazy. 

And then he had been acting fairly good since then. He doesn’t have 
any religion and I’m Catholic. But I said why don’t you go down and 
talk to the priest. Maybe he would help you. 

Here’s another, from the same conversation. 

A: What if you won’t come. I mean how do I- about- Oh, the last time 

he tried to kill me he sat and wrote a long suicide note or whatever. 
I don’t know. I didn’t read it. This was on a Sunday when the kids 
and I got home from church and he wanted to know if I went to 
church with the kids and they said of course. She always goes to 
church with us. He said I know she’s got a boyfriend. I said quit 
acting silly in front of the kids. What’s the matter with you. He says 
oh, and then, I don’t know. Anyway, this time he tried to kill me. He 
wrote this long note. 

—+l just acted like I thought he was kidding. I didn’t want him to think 

Fall 1964, tape 5, side 1 


Lecture 2 


I was taking him seriously. He said well Joey run down to the police 
station before I do something I don’t want to do. I said Daddy quit 
it. Joey says Daddy I don’t want to go down there, they’ll all look 
funny at me when they read the note. I says Joey run outside, Daddy ’s 
only kidding. He says no I’m not. You’d better let me do it. Then he 
got in the car and went tearing off. I looked for the note last night and 
he didn’t have one, so I thought oh maybe he knew I’d wake up and 
maybe not. But I don’t want to leave it go. 

Here’s another, from an altogether different conversation. 

A: I mean the thing that makes it even more serious to me is the once or 

twice that I’ve mentioned it, not deliberately, but kind of slipping, to 
the family or anything like that, they try to make a joke of it, you 

B: Well no, see, here we take all of that seriously. 

A: And believe me it’s no joke because as I say I just don’t feel my life 

is worth anything at this point. 

B: Well we take that very seriously and when someone feels that way we 

try to do whatever we can to try to help them work out of that feeling. 
And we’d like to help you. 

A: Okay, fine then. 

And another conversation. 

A: I want somebody to talk me out of it, I really do. 

B: Uh huh, 

A: But I can’t call any of my friends or anybody, ’cause they’re just going 

to say oh that’s silly or that’s stupid. 

B: Uh huh, 

A: I guess what you really want is someone to say, yes I understand why 

you want to commit suicide, I do believe you. 

Recurrently in these conversations, persons say that when they use the line 
“I’m going to kill myself,” others laugh. And that’s not only by self-report, 
I have things from police reports of suicides where the police then ask persons 
around, “Well, did they ever threaten to kill themselves?” and those persons 
say “Well, he said he was going to kill himself but we just laughed it off.” 
And the question I began to address was, what kind of relationship was there 
between the statement “I’m going to kill myself,” and laughter. How is it 
that laughter would be done there? 

Okay, let’s hold that problem now and turn to another set of materials, via 
which we’ll be trying to see what might be involved in it. I said about the 
opening lines of conversations that they seem to come in pairs. And that one 
person could choose the form of greeting he used. And that if one person 


Part I 

could choose their own they could choose the other’s. Now it seems that there 
is a general class of such kinds of things, which I’m going to call ‘ceremonials.’ 
Other examples are, for example, “How are you feeling?” to which you 
return “Fine.” If one person, then, uses a ceremonial, the other properly 
returns with a ceremonial. 

Let’s look at “How are you feeling?” It’s routinely used between persons 
as either a greeting or greeting substitute. And it’s used between persons who 
needn’t have very much intimacy. But there is a smaller group of persons 
included in the circle of persons who routinely use this object. Call the larger 
group ‘others’ and the smaller, a special class of others. I won’t at this point 
go into describing in detail what the properties are of this special class of 
others. Roughly, they are persons who, if one has a trouble, one turns to them 
for help. Without giving some of the ways we could talk about their relation 
to some ‘one’ - call that one Ego - like, for example, they may be kin, I want 
to approach it in a little different way. One of the ways they stand to each 
other is, if something happens to Ego, then, whoever it is that might be trying 
to discover why that thing happened, could refer for explanations to these 
others. So let’s say they’re ‘causally bound’ to the person who may have 

And that could quite easily make it apparent how it is that if such a one 
is turned to for help, they have a feeling of obligation. They would have a 
feeling of obligation by virtue of the fact that if, let’s say, a suicide occurs, 
then, even if they hadn’t been approached for help in the first place, the 
question would be asked, well what was up with that family that she should 
have killed herself? Many things that might happen to Ego will be causally 
explained by virtue of something that the other did. And if others want to 
avoid that happening to them, then when some Ego turns to them, they feel 
like giving help. And of course the fact that these others walk around with all 
kinds of guilt turns in part on that causal relationship. Now this is among 
laymen; you don’t have to have scientific theories to feel this causal 
involvement. Any layman would ask, if somebody says “My brother killed 
himself,” “Well what’s the matter with the family?” That’s where you would 
look for the source. 

Further, somebody who is not a part of this small group of others can 
become causally involved by virtue of the fact that Ego has asked them for 
help in some way and been turned down. If something then happens to Ego, 
it seems that even if you aren’t one of that small group of others, you know 
about the fact that Ego was troubled, how come you didn’t do anything? So 
knowledge of the trouble is often sufficient to bring one into causal 

Now these people, the whole circle, are going around constantly saying 
“How are you feeling?” Properly, the return is “Fine.” And this can be fairly 
dramatic. I’ve sat around in hospitals, and in a hospital persons who are, say, 
recuperating from serious diseases may be sitting in wheelchairs outside their 
room or in the common room, etc. A doctor walks by a person who looks like 
they’re just about to go, and says “How are you feeling?” and they say 

Lecture 2 


“Fine.” Sometimes, however, a person may take that “How are you feeling?” 
and attempt to use it to present their troubles. And one sort of thing that 
happens in that case is that persons who listen when somebody begins to tell 
them their troubles, talk about themselves routinely as ‘softhearted,’ ‘fools,’ 
and that sort of thing. And when persons talk about themselves as softhearted 
with respect to others, it’s probably something like this that’s happened to 
them: They listen, then they find themselves ‘involved.’ Involved, however, 
without the basic properties that would initiate their relevant obligation, but 
not knowing what to do. And not knowing how to get out, either, because 
they ‘know too much.’ 

On the other hand, the fact that there is that ceremonial relation between 
“How are you feeling?” and “Fine,” may set up the following situation. 
Routinely, if you look at first interviews (and perhaps later interviews also) 
between psychiatrists and patients or possible patients, they begin like this: 

A: How are you feeling? 

B: It’s a long story. 

A: That’s alright, I have time. 

What is this “It’s a long story,” and things like it, doing here? The person 
knows that the line “How are you feeling?” is a ceremonial line, and it’s a 
breach of the proper forms to begin to launch right then and there into what 
it is that’s bothering you. So what they then do is try to initiate another 
ceremony which would provide the basis for them talking. Typically this 
other ceremony is nicely done, in that what one does is offer a tentative refusal, 
like “It’s a long story” or “It’ll take hours,” which turns it back to the other, 
referring to their circumstances; for example, their schedule. And it invites the 
other to then say that their schedule does not control your activities, so go 
ahead and talk. 

Now, persons who are causally bound are obliged to give help when help 
is asked for. That means in part that they’re in bad shape if they don’t give 
help and trouble occurs. They’re responsible for someone. Others hold them 
responsible, and they feel responsible. The question is, is there some way that 
they can go about refusing to give help without ‘refusing,’ in the same way 
that I’ve talked about refusing to give one’s name without ‘refusing’? One 
solution would be to find a way to set up the first remark as the first remark 
of a ceremonial. Because then the proper return is a ceremonial. While there 
are some ceremonials that come off strictly by virtue of the particular object 
that’s used, there are others that are classes of ceremonials. Three common 
classes are jokes, games, performances. They all have the character that the 
next move - or some other given move in the sequence — is the end of it, and 
that’s the end of the whole thing. You tell a joke, there’s a laugh. A game has 
a set beginning and end. A play has the same character. 

That is, I think it’s the fact that we have ceremonial relationships between 
various objects and their proper returns, that sets up the sort of business we 
started off with: “I’m going to kill myself’ followed by laughter. When 


Part 1 

somebody says “I’m going to kill myself,” if the other can cast it into one of 
the ceremonial forms, then that can end the interchange. One wouldn’t 
have heard the ‘cry for help.’ One would have heard a joke. And one would 
have behaved properly with respect to a joke. And it appears that, 
alternatively to giving help, one gets cases of just those three common classes 
of ceremonials. Somebody laughs, or they say “Nice performance,” or “Quit 
playing.” And that would provide, then, for closing that thing off without, 
however, having been in the situation of refusing help in the sense of saying 
“no,” or other such things. So we can see how that form provides for this 
thing to happen. 

We can also see how awfully painful it must be for persons who are deeply 
troubled, and who constantly have people coming up to them and saying 
“How are you feeling?” when they can’t come back. Now and then we see 
that very problem referred to in a joking form. Here is an instance. 

A: How are you feeling? 

B: You really want to know? ha ha 

A: ha ha 

That is, someone, asked “How are you feeling?” jokingly proposes: What if 
I were to take this, not as a ceremonial form, but as a serious invitation. Then 
where would you be? And when people are asked “Well why don’t you tell 
somebody?” they say “It’d be like a melodrama!” or “How can you tell them, 
they’ll just laugh!” 

I want to say another thing about ceremonials. Here is something very nice. 
Very lovely. Lovely in a way, but quite awful, also. When I was thinking 
about this stuff, I came across a very frequently recurring kind of statement. 
I’ll just give one case; a long extract in which a widow is telling a psychiatrist 
of some problems she is having with her married daughter. 

A: Well, I’ll tell you really what got into me last week. You know I was 

just talking about Thanksgiving beingThursday, and she had to 
— ^prepare, but she didn’t invite me. And I go home and I start to think 
about it, and you know, when I spoke to you alone there a couple of 
minutes, I shouldn’t have talked about that, because there was 
something else that was- I mean I touched on it, but there was 
something else. 

- *T just had a feeling that I wasn’t wanted anymore in their house. At 
least by her husband. Naturally she can’t do anything about it. You 
know, I mean if she could, she would start fighting with him, and I 
wouldn’t want to be a cause of that you know. But I thought that 
because, when I first went to the doctor that I went to, this internist 
I was going to last July, and she suggested that I go to a doctor in the 
Valley that she knew. She says well, it’s a good idea because if you 
have to be hospitalized, or if you’re depressed or anything, you could 
stay with me for a couple of days. She says I can’t get down to see you 

Lecture 2 


that often, with the children. But I’ll take care of you if you stay at my 

So this is in July. And I wasn’t able to go to him because I didn’t have 
the money to go. I finally in October, had to go to the hospital. And 
I was there for three days and got these tests, which just made me 
awfully weak, and when I got out she called for me because she had 
my car while I was in the hospital. She called for me and didn’t 
—••ask if I wanted to stay over that night. I get out of the hospital and 
I have no- and I have to drive home, and I felt so weak by myself. 
I mean, she couldn’t because otherwise she couldn’t have gotten back. 
I mean it was just one of those things. But the better thing would have 
been if I could have stayed up there at least overnight and when I felt 
—’•fresh, take the trip down. But she didn’t even ask. And I know it isn’t 
like Lila not to ask, when this was the original reason for my going up 
there. And I just know that she was warned that she better not bring 
me home. 

And of course I started feeling sorry for myself. And then, when we 
were there Wednesday, she said something about preparing a 
—’•Thanksgiving dinner next Thursday and she didn’t say anything 
about me. I figured, well, instead of the family, which we always had, 
the family together, it’s not at my house, it’s at her house. I mean 
during the time I was married I used to have seventeen or twenty for 
dinner because the whole family. And then she had taken over lately. 
So I thought well, maybe she’s gonna have her son. And it’s not up 
to me to expect her to have me. And then I thought well, maybe she 
figured Jay 

that’s the son 

is going to be there, and we’re not getting along right now, and she 
is leaving me out in the cold. And I just began to feel sorry for myself. 

Etcetera. Then she goes on to say: 

Well, it turned out that she said to me, when I said for Thanksgiving, 
— 1 ’•“Well don’t I always have the family?” I said “Well you didn’t ask 
me, how am I supposed to know what’s going to be this year?” I 
mean generally I don’t stand on ceremony, but conditions are, they’ve 
been different lately, you know. 

A recurrent thing that I’ve seen throughout this stuff is persons talking 
about not feeling wanted anymore. The question is, how is that kind of 
feeling provided for in this society? And what would be interesting about it 
would be if we could see some way in which, quote, the structure of society, 
provided for the focussing of kinds of troubles. That’s what I think we can see 


Part I 

with this, 1 if we just consider ceremonials a little further. We can note that 
there are classes of events which, between persons who are not terribly 
intimate, get initiated via ceremonials. “Would you like to come over for 
dinner tonight?” “Sure.” That is, for these kinds of events to occur, there has 
to be an invitation, an offer of some sort. So that’s one task of ceremonials - 
they do the job of providing for these events to take place. 

They do another job, in a way. When persons are quite intimate, then one 
way they measure that is by virtue of the fact that invitations are no longer 
relevant. You can go over to their house without being invited. And people 
will say to each other, “Come over any time you want.” Now with a husband 
and wife, one gets a version of this not feeling wanted, which goes something 
like this: 

Wife : Why don’t you ever ask me to go out to dinner anymore? 

Husband : If you want to go out to dinner why don’t you just say so? 

Wife : I don’t want to go out, I just want you to ask me. 

What she’s picking up here is the absence of ceremonials. And ceremonials 
have this double use. On the one hand they are properly used to provide for 
persons to do things - come over, go out to dinner, etc. - at some state of a 
relationship. At another state those things happen without them. And they’re 
not absent. Indeed, it surely happens that somebody might say, “Well why 
don’t you come over tonight?” and the other says “Why are you suddenly 
making a big deal of it?” But this double use then provides that when 
somebody has some doubts of some sort, they could focus right in there; that 

1 Throughout this volume many of Sacks’ pronominal uses have been changed. 
Here, the operation is more or less innocuous. What is rendered as "... if we could 
see . . .” and “That’s what I think we can see . . .” actually goes, "... if we could 
see . . .” and “That’s what I think you can see,” i.e., the second ‘we’ is actually 
‘you.’ This change instances an editorial policy concerned with solving ‘direct 
address’ as a problem to a reader (e.g., “You ought by this time to be quite aware 
of the fact that . . .”). The policy takes as a resource and license Sacks’ own use of the 
pronouns ‘you’ and ‘we’ in alternation (e.g., “We want to do [X and Y], You want 
a method that generates this.”), and his somewhat eccentric use of ‘we’ in particular. 
For example, he will use ‘we’ when he himself is the obvious referent (e.g., 
“Remember we said about the opening lines of conversations that they seem to come 
in pairs”) or when the class is the obvious referent (e.g., introducing a “much more 
interesting thing that I doubt we’ve noticed”). For a more elaborated discussion, see 
Appendix A. 

Of the range of changes made to the unedited transcripts, very few are marked and 
explicated. It might also be noted that the faithfulness of the unedited transcripts to 
the very words is in principle suspect. Such preservation of the very words as there 
is, is variable. That was not part of the enterprise - with the exception of the 
retranscribed Fall 1964 lectures, which were produced after Sacks’ death. All of 
which is to say that the spontaneous nature of the lectures themselves, the variable 
fidelity of the transcripts, and the manifold changes made in the editing, result in a 
version of Harvey Sacks’ work which from start to finish was in one sense or another 
not under Sacks’ control. 

Lecture 2 


they see this thing is absent, and see the absence via the position of one who 
is not in the position of intimacy. And they don’t know quite how to handle 
that matter. Because if they complain, they get “Why are you standing on 
ceremony?” and if they don’t complain and don’t get the invitations, they 
figure “Jeez, it’s the case that I’m not wanted anymore.” 

Now I can’t make any statement psychiatrically about why persons would 
pick up that double use of ceremonials and use it - or feel used by it - with 
their doubts. But in any event, one can see how it is that the fact that those 
things get used that way, provides a locus for troubles to get focused on. 

Here is another, related, phenomenon. 

A: Hope you have a good time. 

B: Why? 

The “Why?” here is quite apparently a paranoid return, and the whole 
conversation from which this comes makes it quite clear that the person who 
produces it is paranoid. I won’t quote the whole conversation, I want to just 
focus on this interchange for reasons I’ll make clear. 

One of the things that’s reported about persons who have to deal with 
paranoids is that they feel weak, experience a terrific lack of control when they 
encounter them. Now you could go about trying to examine that, perhaps by 
studying let’s say the comparative dynamics of the persons, or various other 
things. But I think you can get an idea of how they would have that feeling 
of weakness by just examining an interchange like this. We’re talking about 
ceremonials. The normal answer to this “Hope you have a good time”is 
“Thank you.” And if one uses a line like “Hope you have a good time” one 
can expect to control the return of the other. In this case the line doesn’t 
control the return of the other, and we can at least begin to see what it means 
to feel weak: Having an expectation of doing something as controlling, and 
finding out that it isn’t that at all. 

But furthermore, this “Why?”-return casts this “Hope you have a good 
time’ ’ into the character of an ‘accountable act. ’ Normally, when one does an 
accountable act, one knows that one is doing an accountable act. This one 
comes off like this: 

A: Hope you have a good time. 

B: Why? 

A: Why? Well, I just would like- you know, you ought to have a good 

time if you’re going on a trip. 

What seems to be involved here is, doing something that wasn’t seen as 
accountable, having it turned into something that is accountable, one doesn’t 
have an account. One offers, then, an account which one feels is quite feeble. 
It’s feeble in a special sense: Not only is it inadequate, but it’s inadequate by 
virtue of the fact that there’s no reason to have had an account in the first 


Part I 

place. But when one delivers the account, one may only see that it’s feeble, 
and get the sense that, “Jesus I’m behaving inadequately here.” 

And that character, that others can by virtue of their return cast your 
activity into something other than it was produced to be - or that they can by 
virtue of their return cast it into what you thought it was - is a very basic 
problem. I call it Job’s Problem. Remember the Book of Job? Job is a rich 
man, doing marvelously. Then everything is destroyed. Job’s position is that 
he didn’t do anything wrong; this was not ‘punishment.’ And now his friends 
come, and they say to him, “Just confess to what you did wrong and 
everything will be fixed up.” That is to say, the appearance of his pain and of 
his loss is sufficient indication for them that he did something wrong. And the 
problem as they see it is that he isn’t about to confess to it. Job, then, is in this 
position of, “Well Christ, the world has changed for me. And maybe I did do 
something wrong.” But he is not about to acknowledge that. But most people 
do. Most people, when they get into a situation, will say, “What did I do 
wrong?” or “What did I do to deserve this?” That is to say, treatments are 
‘proper treatments.’ And one isn’t in a position of saying right off, “You’re 
treating me wrong.” Rather, one finds, the treatment occurred and it must be 
about my action. 

Lecture 3 

The Correction— Invitation Device 

Let me give a quotation. 

A: Do you have a gun at home? 

B : A forty-five 

A: You do have a forty-five. 

B: Uh huh, loaded. 

A: What’s it doing there. Whose is it. 

B: It’s sitting there. 

A: Is it yours? 

B: It’s Dave’s. 

A: It’s your husband’s huh 

B: I know how to shoot it. 

A: He isn’t a police officer. 

B: No. 

A: He just has one. 

B\ Everyone does, don’t they? 

A: You have a forty-five and it’s loaded. 

B: Uh huh, 

A: And I suppose maybe everyone in Burnside Park has one. I don’t 


B: No. But I mean a lot of people have guns. It’s not unusual. 

A: Oh sure. I see. 

The first thing I want us to see in this, I think we have two of, more or less. 
It’s this use of “Is it yours?” and then this one, “He isn’t a police officer.” I 
want to call them, and things like them, ‘correction-invitation devices.’ 

By that I mean: Where one wants to get, from the person one is talking to, 
an account of something - why they did something or why they have 
something - one way you can do it is by saying “Why?” Another way you can 
do it is by asking with the name of the class of things you want. For example, 
a woman is talking to an officer from the juvenile division of the police force. 
Her 14-year-old daughter hasn’t been coming home at night. The woman 
called the police, the police found the daughter, and now they’re talking to 
the woman. And they say, “Have you ever had this kind of trouble with her?” 
That is, ‘this kind of trouble’ is the name of the class. She can then say, “No 

Tape 5, side 2, and a brief extract from Winter 1965, lecture 7 pp. 12-13 
(transcriber unknown). 



Part I 

I haven’t had this kind of trouble,” she can say “Yes” and then give some 
instances, or she can say ‘‘No I’ve had other kinds of trouble.” 

Now it also seems that one can ask for an account by naming, in question 
form, one member of the class, of which the account will be another member. 
For example, “Is it yours?” She doesn’t come back and say just “No,” though 
people sometimes do that. She says “It’s Dave’s.” That is, instead of saying 
“Whose is it?” which he said earlier but didn’t get an answer to, he gives one 
possibility and thereby elicits, as its correction, another; the actual class 

For “Fie isn’t a police officer,” the problem is, how is it that the husband 
happens to have a gun? There are classes of good accounts which would 
explain why somebody has a gun - that is, has a gun properly. One member 
of that class is ‘police officer.’ And what could happen is, if “He isn’t a police 
officer” is an instance of the correction-invitation device, and if the device had 
‘worked,’ then the return would be, “No, he’s a such-and-such,” or “No, we 
have it because ...” Here’s another example. Two persons are talking on the 

A: What do you think was the cause? 

B: It’s a little difficult for me to speak now. 

A: Oh it is. You’re feeling badly yourself? 

B : Oh no. It isn’t it. I’m lacking in privacy. 

A: Oh you’re lacking in privacy. Well, why don’t we arrange to talk 


“You’re feeling badly yourself?” would be one account of how it is that B 
finds it a little difficult to speak now, and the return is the correct account. 

I’ll just mention one way that these things get used, which can get us to one 
basis for their use in the first place. When police interrogate persons, one thing 
they do is, instead of saying “Are you the guy that murdered this fellow?” 
they say “Did you hit him with a tire iron?” And the guy says “No,” and 
then they say “Well what did you hit him with?” where the guy hasn’t 
admitted yet that he did it. And it may be the fact that this form is so 
routinely used elsewhere that permits it to set up the possibility of a trap like 

Now, so far I’ve talked about the construction of these correction- 
invitation devices, and said that it’s based on the fact that, using a range of 
classes, you can refer to one member to get another member. We might also 
be able to say something about the basis for their being used in the first place. 
And at least one basis for that is perhaps something like the following. If you 
say to somebody “Why did you do this?” then what they are being asked to 
present is something they may well know they have to defend. And you set 
up a different situation when what they have to present is something they 
know they have to defend, as compared to setting it up such that you’re 
not asking for an account they have to defend, but you’re ‘inviting a 

Lecture 3 


If these different forms can set up basically different situations, that would 
suggest that we’re looking at extremely powerful matters. I don’t know that 
they’re that powerful, but if they are, that’s a very important thing to know. 
That is, that just by the way you set up the matter, without regard now to the 
consequences in a large sense - as in the murder interrogations - this thing can 

The fact that these things are not only recurrent, but that they do work, 
makes them worth looking at for the following sort of reason. Sociologists 
often talk about something called ‘common knowledge.’ And one question is, 
what is it that common knowledge consists of? One thing it can consist of is 
just lists of items that persons know in common. But there are some things it 
would be nice to know about the phenonenon of common knowledge. One 
of them is what we could call its ‘structural properties’ — and we’ll talk about 
lots of them, I hope. Also, how it is that what persons know ‘in common’ is 
organized. Also, is it the case that the organizational features of what they 
know ‘in common’ are also known? 

So if persons know that there are classes of accounts for some action, the 
question is how do they know those classes? For example, do they know them 
only if you name the class, then they know one or another which are members 
of the class? What this stuff seems to suggest is that on the one hand they 
probably do know, to some extent, the classes and items of these classes by 
virtue of the class name; ‘kinds of troubles,’ for example. But they also know 
them in this fashion: You can name one, and they know, by virtue of the use 
of that one, what class you’re referring to, and can give you another. And 
that’s a non-trivial way of seeing that, and how, common knowledge has its 
organization seen and understood by Members. 

Now I haven’t yet been able to track down when this thing works and 
when it doesn’t work, or what we might say about the circumstances where 
it might be clear that a person knows how to use this, and knows what 
another account is, and doesn’t pick it but instead answers only ‘the question:’ 
“He isn’t a police officer,’’ “No.” 

And in that regard, another question about this organization of common 
knowledge and the members of the class ‘accounts’ would be, how substi- 
tutable are accounts for each other? Is it the case that one is as good as another? 
Which ones would be as good as another for this or that account problem? 
And I ought to mention that the correction— invitation device may not only 
work for accounts, but for all sorts of things; that is, where you can name an 
item, and get in return another item. 

Let me turn to another sort of thing that we can see in this piece of 
conversation. This line, “Everyone does, don’t they?” is one of the most 
fabulous things I’ve ever seen. Where persons are engaged in trying to get an 
account from somebody, there’s an object that the person who’s being 
questioned can slip in. This is one of them. And what it does is, it cuts off the 
basis for the search for an account. I don’t have a terribly elegant name for it. 
What I called it was, ‘account apparently appropriate, negativer.’ Or A3N. 
So, for example, having a gun is something for which an account is apparently 


Part 1 

appropriate. The search goes on for an account, this thing goes in, and now 
an account is no longer to be sought. And this thing isn’t an account of how 
she happens to have a gun. That would be quite a different thing. 

Now these are extraordinarily interesting. One of the most interesting 
things about this one in particular - though it’s not so for all of them - is that 
it seems to be a ‘general purpose A3N.’ By that I mean, it doesn’t much 
matter what it is that you’re seeking an account about, you can use this one, 
‘everybody does.’ This object cuts off accounts about God knows what — 
where accounts are, of course, extremely crucial phenomena. It’s a general 
purpose device. And we’ll see some more later on, some of them much more 
extravagant than this. Just consider, with respect to the organization of the 
social world, that we’re told how fantastically complex it is. How everything 
is a blooming, buzzing confusion. How everybody is different. Etcetera. That 
there are these general purpose devices might give a glimmer, perhaps, of an 
extraordinary kind of simplicity. 

By and large I’ve only talked about verbal interaction. Let me just mention 
something that isn’t a verbal device. My parents live in an ‘exclusive’ suburb. 
And when I was a kid in high school I always used to walk around at night 
in the streets. And when you walk in the streets at night in exclusive suburbs, 
you’re liable to get - as I was routinely - picked up by the cops. “What are 
you doing?” “Just walking.” Then they would take me and stand me in front 
of the police car with the light shining in my face and call up the police station 
to find out if I indeed did live there. This happened night after night. Finally, 
someone gave me the solution. If you bought a dog, that was the end. You 
never got stopped. And that has now become a matter of common 
knowledge. It has become so much a matter of common knowledge that in 
the book Beverly Hills is my Beat by a Captain Anderson, head of the Beverly 
Hills police force, he writes, “It used to be the case that an excuse to walk the 
streets was having a dog. However, the robbers started walking around with 
dogs. But don’t try it in Beverly Hills, because we also know the dogs.” 

So we can begin to locate a range of general purpose A3Ns, with greater 
or lesser generality of application. ‘Everyone does’ has enormous generality. 
Another thing to notice about it is, it doesn’t seem that evidence needs to be 
offered. That is, it’s not the sort of statement about which someone will say, 
“How do you know that’s so?,” where there are lots of statements which will 
get such a question. It’s been known for a long time that there are classes of 
objects - a very predominant class of which are proverbs - about which, on 
the one hand, Members don’t have doubts, and on the other, it’s not a matter 
of evidence that they’re so. And the existence and use of such objects is fairly 
obviously the basis for a great deal of philosophy. Hume, for example, talked 
about the fact that when he was sitting and doing philosophy, there were lots 
of things he could doubt. But he found that as soon as he got up and walked 
out of his study, they were just there. And in an important sense, he had never 
doubted it. It may well be that these are the sorts of things he was trying to 
figure out what in the world they were, and how it is that they seem to do 
what they do. 

Lecture 3 


Again and again we find that when such general-purpose devices as A3Ns 
and proverbs are used, others don’t attempt to question them or contradict 
them. I think there’s some reason why we don’t much see attempts to 
question or contradict these things, and that is that they may be such basic 
objects - that is to say, Members are so committed to their correctness - that 
if you undercut one, exactly what you’ve undercut is not clear. And one 
doesn’t know exactly how we can continue talking. 

My reason for saying that is the following. A woman was collecting 
research materials by going into parks with her children and just starting 
conversations with people. One of the things she reported was how the 
conversations began. And one recurrent way they began was, there would be 
a woman sitting on a bench. This woman would go over to the bench with 
one of her children, and sit down. The little boy would wander around for 
awhile, then he’d come up to her and she’d say, “Go away, I want to sit and 
rest.’’ Sometimes he’d go away, but sometimes he’d sit there, annoyingly. 
And then the other woman would turn to her and say, “They’re all like that, 
aren’t they.” And she’d say “Yeah” and they’d get into a conversation. I 
asked her, “Did you ever say no, or something like that?” And she said 
“Yeah, when I first got out of college I was all full of information. People 
would say that to me and I’d say ‘Well I don’t know, my kids aren’t.’ And 
they’d always stop talking right then and there.” 

I don’t know that that’s generally true. I’d like to see whether it’s so, that 
if you don’t express commitment to these sorts of things, person feel that they 
can’t really talk to you. But apparently proverbs and things like them have to 
be affirmed or membership is not seen as something both of us hold in 
common. That is, these things are known to be so - whatever that means - 
but if you ask for evidence for them, then apparently it’s not clear what kind 
of a box you’re opening up, what sort of things you’re going to ask for 
evidence about next, what would stand as acceptable evidence. So they’re just 
known to be so, whatever that may mean, in that they can be asserted, they 
can be used in conversations, etc. They’re not known by virtue of ever having 
been established in some specificable way. In that sense, they’re strictly 
traditional pieces of information. 

Lecture 4 

An Impromptu Survey of the 

Books like Plainville, Street Comer Society , The Gang, The Irish Countryman, 
and a series of others, were part of a kind of sociology done in the United 
States mainly 25 to 30 years ago. It’s associated with the University of 
Chicago, and also with Harvard. At that time those fellows were trying to 
build ethnographic studies in a tradition that had been developed largely in 
England in social anthropology, and there largely by studying tribal societies. 

That work essentially died out in the United States. But in recent years, 
anthropologists are again returning to detailed ethnographic work, and the 
term ‘ethnographer’ which had fallen into considerable disrepute, has been 
adopted as an ‘in’ term. The Urban Villagers by Gans is one recent book 
which is again attempting to do that sort of work. Two other recent books in 
the same vein are Millways of Kent by John Kenneth Morland and Blackways 
of Kent by Harlan Lewis. 

This recent work is of a new sort, in a way. Where much of the early work 
was criticized as being impressionistic, casual, not hard; that is, not repro- 
ducible, not stating hypotheses, etc., the new ethnographic work - which is 
calling itself things like ‘ethno-cognitive studies,’ ‘ethnocultural studies,’ 
‘ethnoscience’ and the like - is attempting to proceed without being subject 
to those criticisms. The concern is to try to describe the categories that 
members of a society use, but to describe those in a very hard fashion. 

There are several bases for this renaissance - if it’s a renaissance. First, the 
development of very strong tools by linguists. Second, the impact of the work 
of Whorf, whose collected papers are now in a paperback called Language, 
Thought and Reality. Third, and of pretty much equal importance, is 
Wittgenstein’s later philosophy, which the anthropologists who do this work 
are familiar with. The relevant books are two volumes put out as one called 
The Blue and the Brown Books, of which The Brown Book is the easiest 
introduction, though it’s not easy, and the book called Philosophical Investi- 

Tape 3. These materials were probably produced in response to a student’s 

Transcriber unknown. This is the first of several lectures comprised wholely by an 
early transcript for which there is no tape; i.e., which could not be retranscribed. 
There are several of these ‘unknown’ transcribers whose work is included here, 
without whom the first set of lectures would be significantly impoverished. 


Lecture 4 


My own relation to that stuff is fairly tangential in some ways. Instead of 
pushing aside the older ethnographic work in sociology, I would treat it as the 
only work worth criticizing in sociology; where criticizing is giving some 
dignity to something. So, for example, the relevance of the works of the 
Chicago sociologists is that they do contain a lot of information about this and 
that. And this-and-that is what the world is made up of. The difference 
between that work and what I’m trying to do is, I’m trying to develop a 
sociology where the reader has as much information as the author, and can 
reproduce the analysis. If you ever read a biological paper it will say, for 
example, “I used such-and-such which I bought at Joe’s drugstore.” And 
they tell you just what they do, and you can pick it up and see whether it 
holds. You can re-do the observations. Here, I’m showing my materials and 
others can analyze them as well, and it’s much more concrete than the 
Chicago stuff tended to be. 

And I differ from the modern anthropologists, though I would recommend 
that work very much. There is a paper by Hymes called ‘The ethnography of 
speaking’ which is quite interesting. And then there is a collection of these 
anthropologists' most recent works called Contributions to Cultural Anthro- 
pology edited by Ward Goodenough. The trouble with their work is that 
they’re using informants; that is, they’re asking questions of their subjects. 
That means that they’re studying the categories that Members use, to be sure, 
except at this point they are not investigating their categories by attempting 
to find them in the activities in which they’re employed. And that, of course, 
is what I’m attempting to do. 

There are other matters of a deeper sort which are perhaps relevant to why 
sociology took the course it did, and they’re intrinsic in Durkheim’s work. 
One of them is the notion that the order of social events is macroscopic, in the 
sense that you had to assemble lots of events to find statistically what it was 
that was doing the work. I think one can begin to see, in the stuff I’ve been 
talking about, that it may well be that things are very closely ordered. And 
what we have may be something like the following. There may be collections 
of social objects - like “How are you feeling?” - which persons assemble to 
do their activities. And how they assemble those activities is describable with 
respect to any one of them they happen to do. That’s a different kind of order 
to a social world. 

In a way, most of sociology could have been irrelevant, and what I do could 
have been done 50 years ago, 100 years from now, etc. As I said before, it 
stands in close parallel to classical naturalistic biology or zoology. 1 In fact, if 
you want to look at something, Darwin was doing this already. He wrote a 
book, Expression of the Emotions in Man and the Animals , where he collected 
pictures and tried to see if there were, for example, standardized expressions, 
and if so, how did they operate? Not until 50 to 60 years after that book was 
published did people again begin to use photographs. So, for example, 

1 This may be a reference to an introductory lecture, of which we have only a 
paragraph-long precis by an unknown transcriber. 


Part I 

Bateson and Meade began to do that some 25 years ago in Balinese 

But by and large the direct study of humans and their behavior wasn’t 
done. And it probably wasn’t done because nobody believed that it could be 
done, or perhaps because it wasn’t interesting for some reason or another. 
More recently, those who have tried to study it very closely - for example, 
Bales in his laboratory work - have done something exceedingly foolish, I 
think. That is, Bales has the notion that you can categorize it as it comes out, 
so that you sit and watch people as they are talking, and write down categories 
of what they’re doing as they’re doing it. That makes it into some kind of 
trick. There’s no reason to suppose that you should be able to see it right then 
and there. (I find it hard to imagine, for example, that a fellow would stand 
next to an electroencephalograph machine, or any other such machine, and try 
to give you an analysis of the data as it comes out on the tape.) Instead, you 
take these little pieces and you try to collect those that look alike, and it can 
take an awfully long time to understand any given one. 

Another thing that might have been involved was the notion that you 
could tell right off whether something was important. So you would start to 
look at what kings did, or to look at votes, or revolution, for example, because 
those were obviously important. But, for example, the whole of biology has 
been revolutionized by the study of one bacteria, though when that bacteria 
was first being examined, no one had any idea that it would do that work. 
And it’s possible that some object, for example, proverbs, may give an 
enormous understanding of the way humans do things and the kinds of 
objects they use to construct and order their affairs. That has to be seen by 
attempting to analyze the stuff. 

And in that regard, a debt is owed to Freud, who did say “Now let’s treat 
patients as sacred phenomena. ’ ’ That is, something that you would study in 
the sort of way that biblical critics have studied the Bible, where the fact that 
you were looking at one line wouldn’t mean that you could only write a page 
on it. You could write 100 pages. You could spend your life studying it. The 
reason that sociologists haven’t studied a line is that they treat it as something 
very ephemeral, where if you treat it as a machine itself, and as enormously 
recurrent, it has quite a different character. But, for example, the American 
philosopher Meade was a most extraordinary figure who proposed that 
psychology was the study of that which is not available to observation. He 
had an enormous impact on sociology, God knows how or why. It may in 
part have had to do with the notion that sociology studies ‘society,’ which has 
been a very troublesome idea because then you start out by saying, “Well, 
society isn’t observable, but Meade has shown that you can study things 
which aren’t observable. So let’s go study things which aren’t observable, like 

But social activities are observable; you can see them all around you, and 
you can write them down. The tape recorder is important, but a lot of this 
could be done without a tape recorder. If you think you can see it, that means 
we can build an observational study, and we can build a natural study. 

Lecture 4 


Now all of this is background. I don’t want to go through the history of 
sociology and show why one does this or that, because first of all if you want 
to do it seriously, you have to know what kind of work theorizing is, and that 
is an extremely obscure domain if you’re going to take it seriously, at least as 
I take this stuff seriously. I have no idea why sociologists do what they do, and 
I don’t want to get into long arguments about matters which really can’t be 
taken seriously. My arguments can’t be taken seriously, Mills’ arguments 
about the effect of Parson’s proposal to reraise the issue of ‘are ideas important 
and what kind of resources do we have for asking that question’ can’t be 
taken seriously. We can talk about it as philosophers in conversation, but to 
talk about it in any serious way presumes that we have an enormous amount 
of information about how the animal operates, which we don’t have. So the 
more material you have at your command, the more you ought to be able to 
pick up items and see their recurrence and get some idea of what they might 
be doing. But the way to proceed is item by item. 

Q: How does this fit into the general definition of sociology and social 
structure, as structured? 

HS : That’s a good question, for the following reason. What has definition 
got to do with anything? Let’s consider what a definition can do. A definition 
could be an epitaph to be put on a headstone: ‘That’s what this was.’ The 
notion that it’s a control of activity - that is, if you don’t define what you’re 
saying you can’t do anything - is an absurdity. Just consider, for example, the 
fact that biology was said to be ‘science of life.’ Now they find out that maybe 
there wasn’t anything such as that, or that it doesn’t make much difference. 
In any case, it never controls any of their work. So one doesn’t know 
what ‘life’ means until one knows all the things that the biologists try to 
show. And you may find out that it’s not the kind of thing you started out 
with at all. 

There isn’t an answer to what you ask. But I’m not sure what you asked 
had, in any deep sense, a right to be asked. If you take it, as I do, that a 
question has to arise out of something you’re trying to deal with, as compared 
to a way you’ve been taught about questions, then your question might 
not arise. 

It’s a big problem about the University as a scene, that you have almost a 
free right to ask questions. You can turn almost anything into a question, and 
it’s not insane. You learn that much of the knowledge that you’re going to get 
is formable as the answers to questions, because after all you have to be given 
it in such a way that you can answer it if I use it as a question on an exam. 

Now that’s a fantastic constraint on scientific research - that, for example, 
the product of research is subject to being used in a quiz. My own feeling is 
that it was the death of academic psychology that it grew up in a university. 
That implies that they did experiments for which it could be seen from the 
start how the results of those experiments would look as answers to quiz 
questions. If a research couldn’t be farmed as a quiz question eventually, then 
maybe you just didn’t do it. Other fields, like biology and astronomy, 
developed well outside of the universities, so they had lots of materials already 


Part I 

worked up by the time they got into the universities to be taught in this 

And there is, for example, the notion of ‘courses,’ where there’s a 
beginning and an end - an organization. What has that to do with the 
physical or social phenomenon you’re studying? And textbooks. You have to 
have introductory textbooks in sociology or sociologists don’t know anything. 
And the way textbooks teach sociology is quite exquisite. I’ll give a marvelous 
example of how you come to learn sociology. There’s a line in Broom and 
Selznick that goes like this: “Roles are more complex than they appear to be 
at first glance.’’ Now there’s a basic sentence that you know as Members 
without having done any sociology, which goes: “X are more complex than 
they appear to be at first glance.’’ And ‘roles,’ which first of all is a concept 
that couldn’t even be looked at ‘at first glance,’ now becomes something we 
learn via that basic sentence that provides a blank for it. And you know a lot 
about roles, you think. A book like this, built up out of these basic ways that 
you already understand your world as a Member, and simply fits into slots, 
is an especially powerful introductory textbook. But what you’re learning is a 
batch of terms, which you can figure you now know something about, by way 
of what you already know about everything that could fit into that sentence. 
But for us, the understanding and use of objects like ‘X are more complex 
than they appear to be at first glance’ is precisely what we want to be 
studying. It’s not something that we can employ to give us the feeling that we 
understand what is going on in the first place. 

Q: How can you repeat the recognition unless there is some label to 
communicate a definition? 

HS : How do children learn to see ‘another bird’ when they saw a bird once, 
or to re-see a car or a friend? It’s a very obscure question, how it is that persons 
learn to see generalities, or see objects again, or see ‘another’ of an object. I 
really don’t know how they do it, though it’s an important thing to learn how 
they do it. But they do it, and they do it with all sorts of things. You do it 
with verbs, adjectives, sentences; for example, you can see “There’s another 
sentence.” You look at them just as objects. 

Now, what we’re doing is developing another grammar. Right now I’m 
using it with respect to verbal activities, or things like gestures, etc. And in the 
same way you don’t have any trouble seeing a variety of things such as birds, 
or “There’s another verb,” you leam to see these things - at least people come 
back and tell me all the time, “Oh, there’s one of those things you 
mentioned.’ You can see it working, doing the thing it does. We want to 
name these objects and see how they work, as we know how verbs work, and 
that sort of thing. We want to see how activities get assembled, as with a verb 
and a predicate and with whatever else, you assemble a sentence. The category 
that you use to name that activity is given by the Members. They have these 
category names, by and large. So what you’re after is a way of describing the 
activity that they have a name for. 

Ideally, of course, we would be producing formal descriptions, as you 
could give a formal description of how you assemble a sentence. It will not 

Lecture 4 


only handle sentences in general, but it will handle particular sentences. 
Grammar, of course, is the model of closely ordered, routinely observable 
social activities. 

“This is Mr Smith, may I help you?” “Yes, this is Mr Brown” was one 
kind of object and worked one kind of way. The way it works is the essential 
thing about it. There may be a large amount of things that Members can do 
with these objects. Or there may not be a large amount of things because 
someone might do something and people will laugh at them, put them away, 
etc. There may be lots of things that people might never do with these objects. 
What we’re interested in is what do they do with them? Whether that’s 
indefinite, definite, or not, is an empirical question. 

We can say some of the things they may do with them. They may do other 
things, they may not do other things. For example, “This is Mr Smith, may 
I help you?” I have not seen occurring in the middle of a conversation, though 
“Hello” does occur in the middle of a conversation. For example, when you 
say “Wait I have to do something” in the middle of a telephone conversation, 
you may come back and say “Hello” though you’d been talking for ten 
minutes. So the latter has a use that the former doesn’t seem to have. “This 
is Mr Smith, may I help you?” is strictly a greeting. The big thing is to see, 
what are the properties of an object which permit it to do this or that task? For 
example, the way the ‘repetition device’ works. There are special properties 
which provide for its use. 

If you really want me to talk about what sociology ought to be about, or 
what relation any of these things has to what I do, I wouldn’t want to do it 
in class, because that’s like taking a position. These can’t be handled seriously 
unless one takes them as the kinds of issues they are; like take a line out of 
a book and try to see how that fellow came to write that. And who knows? 
At least I don’t know. 

Lecture 5 

Suicide as a Device for Discovering 
if Anybody Cares 

I’ll begin with a quotation. This is a suicidal woman, 40 years old, divorced, 
no children. 

A: Well perhaps you want to tell me uh why you feel like committing 


B: ((sigh)) ((sigh)) Well it’s the same old childish reason that everybody 

wants to commit suicide. 

A: Why is that. 

B: You want to find out if anybody really does care. 

There’s a whole bunch of things that are interesting here, and large collections 
of things we have to do if we were going to be able to generate this 
interchange, most of which I’m not going to consider now. For example, you 
might look at the way this caller sets up giving her answer - by the use of 
“Well it’s the same old childish reason that everybody wants to commit 
suicide’’ - and compare it to the A3N device. 1 That is, the A3N can provide 
that an account need not be produced. The sort of line this woman uses might 
provide that the account she is about to produce is not challengeable, needs 
no defending. 

We might also notice how that’s added to by the use of “you.” That is, 
instead of saying “I want to find out if anybody really does care,” she says 
“You want to find out ...” And those usages, where a person says “you” or 
“one” as a way of stating something that they propose thereby to be a 
generally correct remark, and how they are defended, and what kind of 
attacks they can be subjected to, are something we can watch. And I’ll deal 
with these matters later on. 

I’m now mainly concerned with “. . . if anybody really does care” and not 
the particular objects by which this sequence gets done. But I do want to note 
the fact that this first question, “. . . tell me why you feel like committing 
suicide,” can be asked as a sensible and appropriate question to which there 
is expectably or reasonably an answer - that why you want to commit suicide 
is something that you would have information on, or could propose to know. 

A combination of Fall 1964, tape 4, side 2 and another lecture, ca. Spring 1965 
(‘64— ’65). 

1 See Fall 1964-Spring 1965, lecture 3, pp. 23-4, above. 


Lecture 5 


That the question is askable can be considered this way: Given that there are 
sets of question forms which Members use, one of which is “Why do you 
want to do X?” where ‘X’ is some activity, then, ‘suicide’ being an 
activity-category, just by reference to the relevance of that form for any 
activity, it can be applied to suicide. 

How it is that such a question can expectably or reasonably be answered is 
worth some consideration, since for professionals there are classes of things 
which, if you do them or want to do them, then ipso facto you don’t know 
why. And psychiatrists — and psychoanalysts in particular - take it that a 
person who wants to commit suicide doesn’t know why they want to commit 
suicide, in the sense that the psychiatrist could say why they want to commit 
suicide. (And of course sometimes a person says “I don’t know.”) Now that 
fact doesn’t seem to stand in the way of asking the question. And the issue 
then is, what’s the relevance of that question, and what would happen, insofar 
as persons come to know what it is they didn’t know? That’s Socrates’ classic 
problem; that one thing about knowledge is that you know what you don’t 
know, and to the question “Why?” the answer “I don’t know” is sort of a 
deeper answer; that is, it might have an awareness of the character of this 
knowledge as something only professionals have. 

The notion of ‘opinion’ as contrasted to knowledge (and Plato made a great 
deal of the difference between them) and the sheer introduction of a notion of 
‘opinion,’ provides in part for professionals’ talk to laymen. Because one of the 
characteristics of ‘opinion’ is that it’s something which lay persons are entitled 
to have when they’re not entitled to have knowledge — in the sense that they 
can offer it without ever proposing to have to then defend it. Like they say “My 
feeling is such-and-such on that, but I don’t really know,” as a permissable 
way of talking, where one then doesn’t try to find out what kind of defense you 
have for that statement. So in a way, ‘opinion’ provides for the continuing 
discourse between professionals and laymen. And I presume that it’s a means 
or a mechanism by which not just psychiatrists, but perhaps professionals in 
general can talk to clients — by the notion of the permission that ‘opinion’ gives 
to a person to talk. That is, under the control that one doesn’t really know; 
which is to say, one isn’t entitled to know. And very frequently when you see 
“I don’t know” appended to some statement, that’s what it seems to be doing 
- providing that “I’m not entitled to say this,” that is to say, “I can’t defend 
it professionally,” if it’s a matter of professional information. 

But if it’s the case that there’s going to be discourse between clients and 
professionals, or between the public and professionals, then the fact of a 
distribution of knowledge which provides that professionals know and 
laymen don’t know might seem tremendously interruptive unless you had 
some mediating device, like ‘opinion,’ which would permit laymen to keep 
talking even when they find out that they don’t know. Otherwise they might 
not have any way, for example, of even turning to a professional. 

What I want to focus on is, why is it that suicide seems to be a way to find 
out if anybody does care? The question I asked when I was sitting trying to 
puzzle that out was, what are the available ways in this society for going about 


Part 1 

determining that others care, or that one is relevant to others? What are the 
means available for seeing one’s relevance? 

And while I had that stored at the back of my head, I was reading one of 
the greatest books in the social sciences, Witchcraft, Oracles, and Magic Among 
the Azande by Evans-Pritchard. And some of his observations can begin to 
give us a feel for what such a procedure might look like. Here’s what he 
reports. Whenever anything goes wrong among the Azande - if an Azande 
feels lousy, gets sick, injures himself, is economically in trouble, etc. - he 
engages in the following procedure. He pretty much drops whatever he’s 
doing and goes off into the woods with some oracle procedure. Like, say, one 
oracle procedure is they take a chicken and give it a little poison and ask 
questions to the chicken, which the chicken answers by dying or not dying 
upon being given the poison. 

So the Azande takes a chicken and some poison and goes off into the woods 
with it. And he sits down and makes up a list, essentially composed of his 
neighbors. He considers what his state was before he got ill, and then goes 
through this list of neighbors, considering about each person how he takes it 
they feel about his situation. Are they unhappy that he just got married that 
week, that he just got some wealth, etc.? By going through this procedure he 
then locates some persons who he figures would like to cause him trouble. 
And for each person that he has in this way, he offers a name to the chicken 
and gives it some poison. On some giving of poison the chicken will die. The 
person whose name was offered on that occasion is the person who has done 
him the trouble - caused him to have some illness, caused the rain to fall 
before his crops were in, caused him to have a bad hunting trip, etc., etc. And 
once the one who caused the trouble is found out, there is some procedure for 
getting amends. 

Evans-Pritchard reports that the Azande just love to do this. There is pretty 
much nothing that will stand in the way of them stopping and going off into 
the woods and making up a list and sitting down and considering, for all the 
people around, ‘How are they interested in my good or bad circumstances?’ 
Now, this is one rather nice kind of procedure, which is institutionalized in a 
society, whereby persons can take an occasion and determine for themselves 
properly - that is, there is a proper occasion for doing it - whether anybody 
cares, and what they care. 

Let me make a parenthetical remark about the situation of the Azande as 
compared with this society. One of the things that lies at the basis of the 
availability of that procedure for the Azande, and which is not present in this 
society - and which then provides that we don’t do that in this society — can 
be stated in the following way. The Azande do not have an institutionalized 
notion of chance. Things like falling ill, and most particularly things like 
dying, do not occur by chance for the Azande. There is always somebody 
who’s responsible. And there is a set of procedures, the purpose of which is 
to find out who it is that’s responsible. And these are not random procedures, 
because one has some way of finding out, in the first place, who would be 

Lecture 5 


Now it’s not that the Azande don’t have a good notion of ‘natural causes.’ 
They are perfectly well aware of of the fact that you can get ill from natural 
causes. That doesn’t exclude the fact that there’s somebody interested in those 
natural causes occurring. Evans-Pritchard reports, for example, that some- 
body will stub their toe on a tree and then go off with their chicken. 
Evans-Pritchard says to the guy, “Well after all, you know, it’s your fault. 
You stubbed your toe on the tree.” And the guy says, “I know perfectly well 
that I stubbed my toe on the tree, and that the tree caused that trouble, but 
Eve been through this forest hundreds of times and I never stubbed my toe 
before. There must have been some reason, then, why it happened this time.” 
And that, then, provides for the responsibility. So it’s not a matter of they 
don’t have a good notion of natural causes. It’s that they don’t use a notion 
of chance. 

That being so, you can come to see how rather special it must be for a 
notion of chance to be in fact enforced, and how easy it might be for it to 
break down. Because what a notion of chance involves is that something that 
happens to you is not a matter of inquiry as to how it came about. It just 
happened. You simply don’t investigate why this or that trouble arises, for a 
great many troubles. And that might provide for people to do you ill in more 
or less subtle ways. The notion of chance is a pretty tender one anyway, and 
persons suffering various troubles in our society will often feel that they have 
to shed it and begin to employ, for any given trouble, the question “Who did 
that and why? What do they have against me?” That is to say, they no longer 
feel able to - or they feel compelled not to - use a notion of chance where 
others use it. But in this society it’s not proper, and in fact it’s diagnostically 
significant, if you do not use the notion of chance. By ‘diagnostically 
significant’ I mean persons who do not have a notion of chance are persons 
who have the symptoms of paranoid schizophrenia. When some trouble 
befalls them, they take it that it is some persons who are in the business of 
generating it for them. 

Okay, end of parenthesis. For the Azande, then, there is a device which is 
routinely employable for checking out how it is that others attend to your ill- 
or well-being. Once we have some idea what such a procedure looks like, we 
can begin to consider what are the sorts of things that look like that in this 
society? What are the occasions under which one can make up a list like that 
and just sit down and consider who cares and what do they care? I think you 
can find that there are very, very few. 

One such occasion is the wedding. Before a wedding the parents of the 
bride sit down with a big list and have this enormous ball considering 
“Would this fellow be happy that our daughter is getting married?” “How 
would this guy feel?” Some people give parties, they say, to occasion such a 
device; that is, they say “I just gave a party to see who my friends are.” 

But I take it that the most prominent occasion in, so to speak, a person’s 
life, is right after they die. In this society, on the occasion of death, people 
gather around and talk about how important so-and-so was to our lives, how 
much we cared about him, how much we miss him, what a marvelous guy 

36 Part I 

this was. And that’s what this suicidal woman reports. Later in the call she 

B : And daddy died he won’t suffer anymore now the family won’t be 

aggravated And he’s not here aggravating other people He was 
aggravating everybody before he died and as soon as he died you 
know he wasn’t aggravating anybody anymore so they just said he 
was a great guy 

And anybody who’s ever witnessed that scene has learned what an opportu- 
nity it is. And of course it’s a well-known fantasy, seeing yourself as the one 
who died, getting a chance to get those credits which persons never give you 
and that you can’t yourself collect - that is, for which there’s no occasion to 
collect them. You can see how, for someone in pain, that scene after death — 
which is known to everybody as an occasion for having persons propose that 
they care about somebody - may then come as something exceedingly 
attractive, and ‘the only way.’ And how, then, the ‘attempted suicide’ can be 
the attempt to actualize that scene. 

There are, of course, less dramatic devices for considering somebody’s 
relevance by reference to missing them, or absences. For example, when 
somebody comes back from somewhere, the question is, “Did you miss 
me?” as a way of deciding whether it is that one cares. The question of 
absence and loss, then, seems to be a basic way that one has of dealing with 

Now there are other, more specialized devices for doing a similar task. I’ll 
start considering one of them in a slightly tangential way. One of the things 
I came across several times in the telephone conversations I’ve been analyzing, 
involved a widow or widower who was suicidal. They would say that time 
hangs heavy on their hands and what they find is that “nothing happens.” 
Nothing happens to them. And I wanted to see if there was some way of 
finding out how that comes about; that somebody sees that nothing happens 
to them. 

I also have conversations between young married persons. And one of the 
most exquisite kinds of things that young married persons do with each other 
is, they say things like, “Kennedy was assassinated two weeks after we got 
engaged.” I want to give the name ‘private calendars’ to that sort of talk. And 
I want to note that married couples, each one, by themselves, independently, 
construct these private calendars. And what private calendars do is to provide 
for the locating of, not only events within that relationship, but events of the 
world in general, by reference to the relationship. 

Further, these calendars are ‘causally powerful.’ What I mean by that is, 
there are all kinds of events which can be explained by reference to the 
relationship. There is a generic statement: ‘Because A did X, B did Y,’ where 
one can substitute for A, ‘wife’ and for B, ‘husband,’ and substitute for Y the 
event to be explained, and for X the activity which can explain Y. This 
provides a large class of sensible statements which persons in units like 

Lecture 5 


husband-wife are able to employ. Indeed for many events, such statements 
have to be employed; that is, for many events, such an explanation is the 
only sensible explanation. So it’s often said that while you can give a whole 
list of explanations for why it is that somebody succeeded, in the last analysis 
it’s because of his wife. It’s said without knowing the guy, or knowing 
anything else. 

Another sense in which the private calendar is causally powerful can be seen 
in the paradigmatic statement, “That was before I met you and I was lonely 
then.’’ There is a class of logical statements which the logician Nelson 
Goodman named, and pointed to as creating very basic problems for the 
philosophy and logic of science. He calls them ‘counterfactual conditionals,’ of 
which an example is, I think, “If one had lowered the temperature to 
such-and-such a degree, then the following would have happened,” where 
one hasn’t lowered the temperature and the thing hasn’t happened, but one 
has done something else and something else has happened. Many scientific 
statements are made that way, and Goodman argues that there isn’t currently 
a logic providing for them. But counterfactual conditionals are nonetheless 
routinely used, and they are, nonetheless, enormously powerful. Which 
suggests that perhaps a logic can be invented, or that they’re building on 
something very strong. 

Many uses of the private calendar are such uses. See, one of the problems 
in developing a relationship is finding out that the states of the person you’re 
with are to be accounted for by you , and not by the sheer fact that they’re with 
somebody. That is, they want to be able to say that even if they were with 
somebody before, they would still have been lonely. And that’s one wants to 
do with these private calendars. They’re ways of building up, in deep and 
repetitive ways, the relevance of ‘you.’ And perhaps one of the big things 
about marriage is that that’s just what you’re constantly doing for each other. 
The notion, for example, that marriage is made in heaven, is kind of an 
underpinning to the use of these things. That is, it’s an account that would 
provide the basis for saying “That was before I met you and I was lonely 
then.” Our meeting was virtually guaranteed, and it’s just a matter of, until 
then one drifted, and now it happened. By virtue of this causal structure, of 
course, persons who are members of such units have built-in procedures for 
finding that someone cares. And for a lot of things it’s the only way you can 
find the sense of what’s going on. 

Let me point out something about the private calendar that turns out to be 
rather important. I don’t have a very large set of features of these things, but 
one thing I have found out is that if we compare these private calendars to 
everybody’s calendar, then there’s one striking difference between the two of 
them. And that is, everybody’s calendar has, and private calendars do not 
have, guaranteed continuity. Everybody’s calendar runs on into the indefinite 
future, without regard to anybody in particular being present. Private 
calendars end when ‘we’ end. The end of a relationship, in one way or 
another, can provide that there’s no more events on the private calendar. 

Now then, what we can see the widowed person saying, when they say 


Part I 

“Nothing happens anymore,’’ is that with regard to the private calendar 
whereby events between me and my spouse happen and the value of my life 
is found, no more events can occur on it. You can get, then, a sort of task that 
a therapist, or somebody else to whom one of these persons would turn, might 
have. The task is at least programatically simple, whether it’s easy to do is 
another question. It involves bringing them back to the use of everybody’s 
calendar, whereon events can still occur sensibly in their lives. 

I’ll add another thing, and this is somewhat more conjectural, though not 
strictly made up, and it may be relevant for our materials here. For widowed 
persons, the fact that they’ve had a life with somebody is something that the 
other’s death doesn't take away. And they can say “We had a marvelous 25 
years together,” pointing to all the things we did together, how it is that I was 
happy on this day because of what he was doing, because we were together, 
etc. Now, when persons get divorced, something quite different seems to 
operate. Apparently a divorce can provide for the fact that one can’t even 
retrospectively use the private calendar one had going. The fact of a divorce, 
perhaps with the reconsidering of whether one ever did care, and what after 
all they were doing these last five years that led up to this, seems to involve 
that one can’t then use it for the past that one was ‘together.’ That the woman 
in our materials is divorced may then not only provide that she has no current 
access to the built-in procedures for finding that others care which such a unit 
as husband— wife provides, but also that she is deprived of whatever 
retrospective use she might have had of that unit’s private calendar. 

Via this sort of a sketch we can begin to see where the relevance comes of 
having others care. And that is that the whole class of causal statements that 
are built out of such units as husband-wife and the relationships between 
categories in these units, provide an apparatus in which everybody is supposed 
to be entitled to become a member of such a unit and thereby to have these 
things done for them. And if they don’t become a member, given that they’re 
entitled to become a member, they have a clear way of seeing that something 
is missing. It’s not the easiest thing in the world to find a way to say that 
something is missing. But if you have some objects for which there is no rule 
of exclusion in the first place - everybody is entitled to them - then if someone 
doesn’t have it at some point that one is entitled properly to have it, one can 
say that it, and its consequences, are missing. 

We can tie this up to some extent by asking what, then, is the consequence 
of not having persons care? Well, these lay theories — and all these causal 
statements and entitlement propositions are lay theories - have a rather 
interesting property. If you consider our prototype of a scientific theory, then, 
if some object doesn’t conform to what the theory proposes about the object, 
then the theory has to be revised. This world has been constructed in a rather 
more exquisite way. What goes on is the following. A large class of lay theories 
are properly called ‘programmatic theories. ’ If they don’t describe your circum- 
stances then it’s up to you to change. And if they don’t provide for you as a 
Member, then it’s up to you to rid yourself of being a Member, for example 
to kill yourself. In that way you keep the theories going as descriptive. 

Lecture 5 


If you’re a member of one of these units you have essentially automatic 
ways of finding that others care. It’s built into the structure of ordinary 
discourse, and the way persons see how events come off. If you’re not a 
member of such a unit, it’s still relevant, but its structure is not available to 
you. And you may then try that procedure which works for everybody - 
dying - either as a way to find that somebody does care, or as a way of 
providing that the theory that people ought to care is made correct by virtue 
of your no longer being a Member. And we’ll see constantly that persons talk 
of a whole range of things where if something is not so for them, then that 
doesn’t provide that what’s supposed to be so is thereby wrong, but that 
they’re wrong. 

Let me add one more device relevant to “Does anybody care?” It is, of all 
things, trash mail. The next time that they have hearings about removing 
trash mail, I’m prepared to go and testify against its removal. Because trash 
mail is a most interesting thing. I’ve mentioned this woman who used to go 
to the park and sit and talk to people. Many of those were old ladies. They 
were all utterly isolated. They came to Los Angeles after their whole family 
died, or they came with their husband and he died. They live in apartments 
near a park and they spend their day in the park. But they regulate their lives 
in most interesting kinds of ways. 

Even though they have almost no money they, for example, never purchase 
at supermarkets and never purchase more than a day’s food. Because if 
they did, they’d have nothing to do the next day. And they routinely will get 
up - you’ll be sitting in the park talking to them, the only person who’s 
talked to them since God knows when - they nevertheless get up and say “It’s 
11 o’clock, I have to go home and check the mail.” Now there’s nobody 
who’s writing to them. What it is, is that there’s that trash mail coming, and 
that’s something. 

Consider their situation: The mailman comes every day, and they know it. 
And that means that for them, they have to go check the mail every day. The 
only mail they do get is this mail that everybody gets. But for them, it’s 
something. And if they had to recognize that he would come every day, and 
every day they would find no mail, and they could look forward to that day 
after day, then that situation of theirs, of isolation, would so be built into their 
circumstances and shown to them routinely, that it might become far more 
unbearable than it is - and it’s pretty unbearable - because this is a device that 
happens every day, for whomsoever. You don’t know who is getting 
telephone calls, you don’t know how many phone calls are being made, but 
every day, everybody has the mailman go by. And if you just consider the 
comparative cost of trash mail versus an enormous mental health operation, 
then trash mail is not expensive. And for these people it’s by and large the 
only means by which the routinely-used device of delivering mail does not 
become the kind of thing it would otherwise become - this persistent 
statement to them that nobody cares. 

Lecture 6 

The MIR Membership Categorization 


I’ll begin now talking about some very central machinery of social organiza- 
tion. Let me indicate how I came by the findings I’m going to present. In 
dealing with first conversations I’ve very frequently found, as anyone can 
easily find, that especially in the early parts of these conversations certain 
questions are prominent; questions like “What do you do?” “Where are you 
from?” etc. I wanted to see if there was some simple way that I could describe 
the items that those questions contain, so as to provide for their occurrence by 
rather abstract descriptions. That was the initial task. Its consequences are 
rather powerful, and I’ll develop them as I go along. 

It seems that there is a class of category sets. By ‘category sets’ I means just 
that: A set which is made up of a group of categories. There are more than 
one set, each of which can be named, and they have common properties. And 
that is what I mean by referring to them as a ‘class.’ 

A first thing we can say about this class of category sets is that its sets are 
‘which’-type sets. By that I mean that whatever number of categories a set 
contains, and without regard to the addition or subtraction of categories for 
that set, each set’s categories classify a population. Now, I haven’t made up 
these categories, they’re Members’ categories. The names of the sets would be 
things like sex, age, race, religion, perhaps occupation. And in each set are 
categories which can classify any member of the population. I call them 
‘which’-type sets because questions about any one of these can be formulated 
as, “Which, for some set, are you?,” and “None” is not a presumptive 
member of any of the categories. And that would suggest what it is that 
provides for such questions occurring at the early part of first conversations: 
You don’t have to know anything about somebody to be able to formulate a 
set of questions for which “None” is not an expectable answer. And of course 
for some of the sets you don’t have to ask the question. 

A second thing we can say about this class of category sets is that its 
categories are what we can call ‘inference rich.’ By that I mean, a great deal 
of the knowledge that members of a society have about the society is stored 
in terms of these categories. And by ‘stored in terms of I mean that much 
knowledge has some category term from this class as its subject. And the 

A combination of Fall 1964, Ml, sides 1 and 2, and Winter 1965, lecture (4-5) 
(transcriber unknown). The discussions are the same in both sets but are more formal 
in the latter, and much of what is shown here is taken from the more formal version. 


Lecture 6 


inference-rich character of these categories constitutes another warrant for 
their occurrence in early parts of first conversations: When you get some 
category as an answer to a ‘which’-type question, you can feel that you know 
a great deal about the person, and can readily formulate topics of conversation 
based on the knowledge stored in terms of that category. 

A third feature is that any member of any category is presumptively a 
representative of that category for the purpose of use of whatever knowledge 
is stored by reference to that category. So, for example, a foreigner comes to 
the United States and you find yourself asking them about the political 
situation in Ghana, or how they like the food in the United States, without 
reference to whether they stand as a member of the Gourmet Club of France, 
or don’t ever eat out, or aren’t interested much in food, or are just ordinary 
citizens, so to speak. But one finds that it’s done. And it’s done for any of 
these category sets. 

Let me emphasize that we’re dealing with categories, and not necessarily 
with what sociologists might call ‘groups,’ ‘organized groups,’ ‘organiza- 
tions.’ It’s quite important to see that presumptive representativeness holds 
whether or not the members of that category are or are not organized. If they 
are organized, it holds whether or not they choose their representatives. The 
fact that they are organized and choose their representatives does not mean 
that one cannot apply the knowledge stored about such a category to persons 
who are not selected by the group. And, furthermore, the fact that it is not 
a group in the sense of being organized, doesn’t mean that one cannot apply 
such knowledge. 

I’m calling this whole apparatus the MIR device. And that is an acronym. 
‘M’ stands for membership. T stands for inference-rich, and ‘R’ stands for 
representative. That’s the core of the machinery. I take it one can readily 
notice how absolutely central this is, for a vast amount of stuff is handled by 
Members in terms of the categories that it locates and the way it locates them, 
and the activities that those categories are used to handle. 

Now one might get a sense here of a certain problem, and I’ll mention it 
right now. I take it to be a central problem of sociology, and I’ll try to show 
some sorts of solutions to it eventually. The problem is this: There are these 
category sets. For any person being talked of, how is it that Members go about 
selecting the set in terms of whose categories that person is going to be talked 
of? It’s perfectly obvious that there is a range of sets whose categories could 
be used; from the set ‘sex,’ “a woman”. From ‘race,’ ‘‘a Negro.” From 
‘religion,’ ‘‘a Catholic.” From ‘occupation,’ “a psychiatric social worker,” 
etc., etc. Each of these categories could apply to the same person. And it’s 
perfectly obvious that Members do use one set’s categories for some 
statements and another set’s categories for other statements. If we’re going to 
describe Members’ activities, and the way they produce activities and see 
activities and organize their knowledge about them, then we’re going to have 
to find out how they go about choosing among the available sets of categories 
for grasping some event. 

All the sociology we read is unanalytic, in the sense that they simply put 


Part I 

some category in. They may make sense to us in doing that, but they’re doing 
it simply as another Member. They haven’t described the phenomena they’re 
seeking to describe - or that they ought to be seeking to describe. What they 
need to do is give us some procedure for choosing that category which is used 
to present some piece of information. And that brings us back to the question, 
are there procedures that Members have for selecting categories? One of my 
aims is to show that there are. 

For now, let me show one of the tasks this MIR device, in combination 
with a particular sort of operation, can be involved in. I’ll be talking now 
about some extremely basic and extremely generic social control devices. The 
particular sort of operation consists of one way that Members go about 
making new knowledge. Suppose some event occurs and is known about by 
reference to the name of the person who did it. The way you get a piece of 
knowledge involves pulling out the name and putting in some category. Then 
one gets, not ‘John did X,’ but ‘a such-and-such did X.’ In that way one gets 
additions to any given body of knowledge about such categories. And what 
we find is that an enormous amount of what we could call the lay theories of 
social actions are fitted onto these categories. 

Given the MIR device, and given this operation whereby new knowledge 
is formulated - by replacing a name with a category - we can begin to see 
how a class of social control devices gets set up and is used. It has as its basis 
that members of the society are constantly engaged in monitoring events; on 
the one hand by reference to whether something that has happened is 
something that they’re accountable for, and on the other hand, to find out 
what is getting done by members of any of the other categories. 

Apart from the routine monitoring terms of these categories, we get nicely 
special kinds of occurrences which provide a beautiful view of tensions arising 
as persons await the discovery of which of them is going to be found to have 
done this thing. For example, the hours between the assassination of President 
Kennedy and the determination of who it was, and thus what category it was 
that performed the act. If you have access to a variety of materials from that 
time, you can see persons reporting themselves going through “Was it one of 
us right-wing Republicians?,’’ “Was it one of us Negroes?,” “Was it a Jew?.” 
etc. That is, “Was it me?” in that sense. 

This sort of monitoring makes for great sophistication in kinds of ways 
of doing trouble. For example, in the recent Russian economic trials, it was 
quite sufficient for those who were encouraging anti-semitism in Russia, to 
simply publish the names of the persons who were tried. The names turned 
out to be seeable as belonging to Jews. And you could leave the rest to 
everybody’s routine procedures: “See? Jews are economic criminals, as 
everybody knew.” 

What you get then is a whole series of teachings, all of which have the same 
form: “Remember you’re a such-and-such” (a lady, an American, a Negro, 
a Catholic, etc.). That is, any action you take is exemplary. Any action you 
take is something we’re going to have to come to terms with. Such teachings 
belong to a class of activities which are often called ‘internal systems of social 

Lecture 6 


control.’ They may be informal or formal, but they’re not like such things as, 
for example, a government. What they have in common is that they are all 
operated by and enforced by and taught by and used by members of the 
category whose members are to be controlled. 

These internal control devices all seem to be built on, and have their power 
by virtue of, this very simple apparatus the MIR device, which is utterly 
disjunctive to whatever these groups happen to be, or whether they happen to 
be ‘groups’ in the organizational sense of the term. Indeed, some of these 
classes’ categories may set out to become organizations by virtue of the uses 
that are being made in any case of their categorial membership. It is no mere 
plea, for example, that women have got to get organized. If an event occurs 
where what somebody does is seen by reference to the category ‘female,’ then 
women in the society find themselves constructing explanations for how it 
would have happened, or proposing that they can’t figure out how such a one 
would have done that - though of course they need know nothing about the 
person except that category of membership. 

To get some contrast on these control devices, we might notice that they’re 
much unlike a characteristic thing to which we give the name ‘scapegoating.’ 
In scapegoating, if a member of a group does something which is sanction- 
able, then either that member or some other member of the group may be 
selected by some procedure - perhaps the group itself selects them, perhaps 
an outsider selects them - and a sanction is applied to that person or persons. 
That being done, guilt is purged. This doesn’t happen. If somebody does 
something that is formulated as an action of the group, then a piece of 
knowledge is thereby legitimately usable and you don’t have some procedure 
for cutting off its use - such as the scapegoating procedure - where it’s 
thereafter not proper to say about a group that they did this thing. So, 
Members’ stake in the actions of other Members is not conditional in that 
sense; that is, it is not conditional on being or not being purged. 

Now, given what I’ve said so far, there are certain things that can be 
examined pretty clearly. One of them has to do with a situation where some 
person who is a member of some category does something to another person, 
where that other person happens to be a member of the same category. The 
pain of it, that it’s something awful, operates not merely by virtue of the fact 
that somebody did something to you for which you have some legitimate 
feeling of being injured, but in addition you have the sense of having injured 
yourself. And in such cases people talk of being ashamed of being a member 
of the category they’re a member of. 

So, for example, I have materials in which a person in one or another 
trouble will call the social agency of their religious denomination (Catholic 
Family Service, Jewish Welfare Group, etc.), and get rejected. I have two 
cases in which such a person subsequently calls someone else and is reporting 
this. And in each case they report - as is not reported when persons are 
rejected by, say, a municipal agency or some hospital - “It’s not just that I 
was rejected by somebody, but I feel ashamed of belonging to a group that 
does that to people.’’ That is, impersonally they are now observing the group 


Part I 

in operation, where they take it that whoever it was that spoke to them is a 
representative of the category (Catholic, Jew, etc.), and that such a one did 
something to another is something that reflects on themselves and devalues 
their membership. That is, not only non-members, but members of a category 
take it that the actions of that category can be assessed. It’s not merely that 
a non-Catholic could hold this up and say, “See? Catholics don’t take care of 
their own,” but that a Catholic will say, about their own group, the same 
thing. The generic importance of such a phenomenon is that it’s not just one 
category’s view of another, but that knowledge is standardized across the 

I want now to notice several affiliated features of the MIR device. I’ll start 
off with a feature I found in a few pieces of conversation - and indeed it was 
looking at these materials that really launched a lot of the considerations here. 
First I’ll show the materials. In the first extract, two teenagers, a fellow and 
a girl, are talking about dating. 

A: Corliss the g- this chick that I’m hanging around with now, she’s real 

nice she’s got a real good personality, she’s not- you know she’s // 
just a real cute little kid. 

B: Mm hm, 

A: And last night we went to the Mardi Gras together and we were both 

well we were both pooped because I-I ran in the track meet yesterday. 
And she-she’s in the girls’ tumbling team. I mean she doesn’t like it 
she’s just on it because she needs the credits. 

The second is from the emergency psychiatric hospital. 

A: How old are you Mr Bergstein? 

B: I’m 48, I look much younger. I look about 35, and I’m quite 

ambitious and quite idealistic and very inventive and conscientious 
and responsible. 

What struck me was that a thing was presented, “She’s on the girls’ tumbling 
team” and “I’m 48,” and then you get a modifier in a long or short 
statement. At this point in our considerations I take it that what we have here 
is very apparent. For some category, like these two, there is a set of things 
known about a member which can be applied to any member, for example 
that being on the girls’ tumbling team is presumably something very gauche, 
and that someone who is 48 is past their prime, which any person now talking 
about such a category membership has to come to terms with. 

So we have a class of things, these modifiers. And they consist of attempts 
to provide that what it is that may be said about any member is not to be said 
about the member at hand. “I’m 48 but I look and feel younger,” “She’s on 
the girl’s tumbling team but . . “He’s a Negro but . . . ,” etc. Having 
seen this, I think we can see a solution to something I had been puzzled about 
for quite awhile. It is, I suppose, a rather minor kind of interest, but it’s 

Lecture 6 


related to just this business. There is a class of what look like tautological 
statements, “Women will be women,’’ “White folks is white folks,” a whole 
large bunch of them. It looks like they are simply tautological statements; that 
is, that they say nothing. But given the use of MIR modifiers, these 
tautological statements seem to be - to use an awkward phrase - ‘anti- 
modifier modifiers.’ That is to say, under a condition where for some reason 
it’s proposed, or one has been going along under the notion that, the person 
whose behavior is being considered is to be classified by reference to one of 
these modifiers - for example, “He’s a Negro, but the things you can say 
about Negroes you can’t say about him” - you have this other class of 
statements available to flip in and provide that in the last analysis he’s like the 
others. They provide for the re-relevance of whatever it is that’s known about 
the category. And if you watch conversations in which these things occur, 
that’s the way they get used. 

In certain kinds of relationships these anti-modifier modifiers can be 
extremely deadly kinds of things to have around, because they can always be 
used in the following way. There is the kind of relationship that proceeds 
between some set of persons under the modification that what is known about 
the category that one of them is a member of, is not to be applied to this one. 
And now the one who is living under that modification has always to carry the 
possibility that some time somebody is going to say, “Well, now it really 
comes out ...” and invoke a set of things that can properly be said about that 
category, removing the modification under which they’ve been living. 

There is another feature I want to mention. Some of the category sets of the 
MIR device have to be differentiated from others in some special ways. If one 
considers categories like age and social class, in contrast to those like race and 
sex, one finds some rather interesting differences. For all of them there may 
well be a stable set of categories used by everybody. But whereas for, say, sex 
and race it will be by and large the case that one can take it that whatever 
category somebody applies to somebody else or to themselves, anybody else 
would apply that category, that is not so for categories like age and social 
class. What you have with these latter sorts of categories is a rather lovely 
series of things going on. If any Member hears another categorize someone 
else or themselves on one of these items, then the way the Member hearing 
this decides what category is appropriate, is by themselves categorizing the 
categorizer according to the same set of categories. So, if you hear B categorize 
C as ‘old,’ then you would categorize B to decide how you would categorize 
C. And again, the same procedure works for such a thing as social class. 

This sort of operation is probably basic to something which sociologists are 
talking about as a generic matter (but which is by no means generic) and which 
Members also use. And that’s the notion of ‘perspective.’ If somebody calls 
somebody ‘old,’ what you want to know is, how would you call the person who 
called the person old? That way, even if you don’t know the person being 
categorized, you can have, for example, some notion as to the range that’s 
involved. If you’re an adult and it’s a ten-year-old who calls somebody old, 
then you can figure that the somebody could be anywhere from 20 on up; 


Part I 

that is, they’re possibly young. Now it may well be, though I can’t say this 
with any confidence, that even though this sort of business is going on, it’s 
largely the case that the same information is stored for any category. So that 
A and B, being from different categories, may place C in different categories 
(A categorizing C as old, B categorizing C as young). But whoever A and B 
would place in the same category (whoever A might call old and whoever B 
might call old), they would say the same things about that one. So that when 
kids talk about ‘old people,’ though they’re talking about somebody who is 
20, they may have the same information about that one that a 30-year-old 
has about someone who is 70. 

It may not be the case, then, that on the different ‘perspectives’ - that is, 
on the different uses of the categories - depends a whole different body of 
information, but that the knowledge is stable for any category like ‘old.’ It’s 
just a matter of what category is using the term ‘old.’ The same thing may 
work for social class. For these, there may be no position that provides for the 
definite classification of somebody. One wouldn’t then find that somebody 
carries around an identity which is stable for any environment they come into; 
for example, that they would be ‘old’ no matter who it is that’s around them. 
Nor would it be the case that for each of the persons around them they would 
be seen as the same person that they have to see themselves as. There is no 
supposition of agreement on any categorization for such persons all catego- 
rizing each other. 

Let me mention another minor thing which this machinery can clarify. 
Where it might not be proper to say certain things about another person, or 
for that matter about oneself, what one can do is to propose membership in 
some category, where that category stands as the adequate basis for inferring 
those certain other things. So, to take a sort of extreme example, here is a first 
conversation between a psychiatric social worker and a suicidal man. Earlier 
the man had said that he’s been married three times, he’s not working and not 
married now. 

A: Is there anything you can stay interested in? 

B: No, not really. 

A: What interests did you have before? 

B: I was a hair stylist at one time, I did some fashions now and then, 

things like that. 

They go on for a couple of minutes. Then: 

A: Have you been having some sexual problems? 

B: All my life. 

A: Uh huh. Yeah. 

B: Naturally. You probably suspect, as far as the hair stylist and, uh, 

either one way or the other, they’re straight or homosexual, something 
like that. 

Lecture 6 


In this case, while it might not be proper for this man to say about himself 
that he’s troubled by possible homosexual tendencies, he finds a way to 
invoke a subset of occupational categories, “hair stylist. . .fashions. . .and 
things like that,’’ which constitutes an adequate basis for inferring homosex- 
uality. And in his subsequent talk he proposes that such an inference has 
“probably’’ been made by the other. Apparently, then, there are ways of 
introducing a piece of information and testing out whether it will be 
acceptable, which don’t involve saying it. 

Now, I have by and large been talking about negative information stored 
in these categories, but they obviously provide that system of rewards which 
any young person can expect by virtue of becoming a member of any category 
that they can become a member of — for example, occupation, or changes of 
religion - since to become a member is to make stateable about yourself any 
of the things that are stateable about a member. So they’re the basic system 
of incentive for persons to do a variety of things. That’s now fairly well 
known, so all the various occupations are engaged in trying to sharpen up 
their images so as to make it attractive for persons to come into them. 

Also, there are paths that can be constructed in terms of set movement, 
which persons may use to consider what their likelihoods are of ever getting 
to be a such-and-such, given that I’m a this-and-that. And the fact that 
‘somebody has made it’ provides a path. But this can be used in very negative 
ways just as well. For example, I heard on the radio the other night an 
interview with congressman Billy Mills. And he was asked, “Well, you’re a 
Negro, you came from the lowest part of the economic structure and 
nonetheless you’ve made a great success, why can’t any other Negro do it?’ ’ 
That is, there is a path which your history has laid out. Why does that not 
stand as a way to get from one place to another which any person in your 
initial state can use? And now Mills has to come up with some argument 
about why it couldn’t or shouldn’t. Or, for example, the notion of ‘token 
integration’ can be used in business establishments in just that way. The fact 
that there is a such-and-such at this place provides that there is indeed a path 
from the position they started with. And since there is a path, and since 
Members at that initial position know how to use it, the burden now lies on 
them to do so. 

Finally, let me offer something to consider. I have no idea whether it’s so. 
It sounds altogether too smooth to me, and nonetheless it also looks, on the 
face of it, to be very descriptive. Many of these classes are, or can be built as, 
two-set classes. Sex is a two-set class. Race can be formulated as a two-set 
class; for example, non-whites and whites. And there’s rich and poor, old and 
young, etc. The question I’m asking is, does it matter, for the kinds of things 
that can be done with these classes, how many sets they contain? Two-set 
classes would seem to have certain kinds of attractions. For example, they’re 
tremendously easy to compare. With a two-set class you can apparently make 
an observation of comparative lack much more easily than otherwise. And I 
wonder, for example, whether many kinds of conflict and perhaps most sorts 
of revolutions occur by virtue of these two-set classes; as we say, the haves and 


Part I 

the don’t haves. Under such a view, you can see all sorts of different things 
being fitted to the notion of haves and don’t haves. Marx can be seen to have 
used this two-set class. The movement for equality of women can be seen to 
be using it. And the Negro revolution as well. 

To establish a two-set class you might start with one group who you locate 
as the group in power, or the haves. Give them a name: Whites, men, the old. 
And then assimilate all the others to some predominant feature of those 
others; for example, a lot of them are Negro so you call it ‘the black 
revolution.’ But if you just go through the ways revolutions tend to organize 
themselves, or the ways movements tend to organize themselves, or notice 
that games - which are model conflict situations - are so often either two- 
party systems or variants of two-party systems, it begins to look as though 
formulating in terms of two-class sets is a method of doing things. Whether 
this is so or not, I haven’t the vaguest idea and it needn’t be taken with any 
seriousness. But what I said earlier, about this device as a basic mechanism of 
social control is, I think, important. If we can come to see what’s involved in 
it, I think we can see something useful, and something that’s of theoretical 
importance in sociology. 

Lecture 7 
On Questions 

A few days ago I asked people to go out and collect first lines of ‘pickups.’ I 
did that because I want to return to some considerations of rules of 
conversational sequencing, and to introduce a rule. But it has initially such a 
vacuous sound to it that I wanted to see if there was some way of making its 
relevance clear, more quickly and more intimately than the materials I had on 
hand permitted. Looking at my materials, these long collections of talk, and 
trying to get an abstract rule that would generate, not the particular things 
that are said, but let’s say the sequences, or their continuity and things like 
that, then you come up with a rule that says something like: A person who asks 
a question has a right to talk again , after the other talks. 

That sounds enormously empty. One reason it sounds empty is, at this 
point I can’t put in “. . . after the other answers.” I can’t put it in, in part 
because ‘questions’ and ‘answers’ are in some ways altogether different 
objects. For example, a question is a grammatical matter and an answer isn’t. 
I don’t think you can locate ‘an answer’ grammatically. Also, a question can 
be paralinguistically described. An answer doesn’t have that character. That 
is, if you play a tape or listen to a conversation and forget about what’s being 
said but just try to get the tone, pitch and that sort of thing, then you can 
describe what a question looks like. A question has a form. And an answer 
doesn’t, apparently. So we can talk about ‘asking questions’ and identify some 
object as ‘a question,’ but we can’t do that very much with ‘an answer.’ 

But anyway, there looks to be a rule that a person who asks a question has 
a right to talk again afterwards. And that rule can provide a simple way of 
generating enormous masses of sequences of talk: Question, talk, question, 
talk, etc., etc. We can say it’s a rule with a repeat device. But what else can 
we get from it? About 60 first lines of ‘pickups’ were handed in. Seven of 
them were other than questions. More than 50 were questions. And by 
reference to this rule, that wouldn’t be an incidental fact. You might begin to 
see why questions get used. 

Here’s a classic Yiddish joke on just this issue. A young man gets on a train 
and sits down next to an older man. The younger one asks, “Can you tell me 
the time?” and the older man says “No.” “What do you mean no?” the 
younger one says. The older one says, “If I tell you the time we will have to 
get into a conversation. You’ll ask me where I’m going. I’ll ask you where 
you’re going. It will turn out we’re going to the same place. I’ll have to invite 

Fall 1964, M2, side 2 and ‘(Fall 1964)’, pp. 1-9 with a brief extract from 
Fall 1964, tape 7, side 1, p. 3. 



Part I 

you for dinner. I have a young marriageable daughter, and I don’t want my 
daughter to marry someone who doesn’t wear a watch.” 

I’ll make something of an excursion and talk a little bit about pickups. Let 
me sort of mock one through; one that I saw at the airport. A bunch of about 
20 people are standing around waiting for a plane to arrive. At the edge of 
the crowd a girl is standing. A guy comes up somewhat behind her and says, 
‘‘What time does the plane arrive?” She turns and says ‘‘In 20 minutes” and 
turns back. Then he asks another question. She turns, answers, turns back. 
This goes on for five or six questions. Then she just turns her body to him, 
without especially any expression, or even looking at him. He keeps asking 
questions. She keeps answering them. She turns her head up when she 
answers and then brings it back. So it looks like if at any point he was to stop, 
that would be that. At some point she takes out a cigarette and he lights it. 

Now, it’s a very well known fact about homosexual society that in bars 
frequented by male homosexuals the bartender keeps a pack of matches 
behind the bar because it’s impossible to ask for a light. That is to say, asking 
for a light is so much a ‘move’ that if what you want is just a light, you just 
pretty much can’t do it. It also seems that between males and females, asking 
for a light is a tremendously sexual thing to do. In any case, in such things as 
pickups it’s a key point. And I don’t think that’s incidental. There are a series 
of areas which are closely regulated, especially between persons who don’t 
know each other, which lighting a cigarette can be involved in. 

First is eye-to-eye contact. If, for example, you sit in the library and have 
somebody sit opposite you watching you, you can get the following results. 
You start to look at somebody. You’ll find that even when you can’t tell that 
the other person has caught your glance, the person sitting opposite you can 
- and they’re not looking at the other, they’re just looking at you. Your eyes 
flick, like magic, across the room. A glance just drives them away. So 
eye-to-eye contact is a highly regulated thing. But of course when a person 
lights a cigarette for you, that’s an occasion for persons to align with heads 
very close to each other, looking at each other directly. 

A second thing is touching; a very closely regulated matter as between 
persons who don’t know each other, and even between persons who know 
each other. But again, lighting a cigarette can be an occasion for taking 
someone’s hand, for example. In any case, some touching can very easily go on. 

It’s also a point when business is being done between the persons such that 
they’re not talking and nonetheless together. Whereas previously, every time 
they stop talking there’s an issue about whether they are together. But here 
you have a first occasion when the persons can stand together, looking at each 
other perhaps, without ever saying anything, and pick it up again. And the 
development of the ability to have a pause in a relationship is very crucial. 
Just remember, for example, when we were very young, going out on dates 
for the first times, how it was that gaps in conversation were treated as such 
tremendously painful things. The developing of an ability in a relationship to 
be silent is a very important thing. And this is a first step to that. 

Another thing this act of getting a cigarette lighted does - this is of course 

Lecture 7 


completely apart from anything like psychoanalytic theories about cigarettes 

- is that when a woman has a cigarette, she has an occasion to be moving. 
You don’t smoke a cigarette with your hands at your sides. And, for example, 
as the cigarette is lighted for her, she can bend her head and look up through 
the top of her eyes - which is in our society a very sexual kind of way to be 
looking at somebody. And when she has a cigarette in her hand, she has then 
a routine opportunity for bringing her hand to her face, which provides the 
minimal condition for him to be looking at her face; that is, it’s perfectly 
legitimate to follow a moving object, your eyes do it normally. So she can 
begin to get into physical action, moving her head and her arms, etc., which 
then provides for, literally, a dance. 

That ends the excursion. I wanted to give some idea about the relevance of 
this rule that if you ask a question you have a chance to talk again afterwards. 

Now, there are some situations which have their particular character by 
way of the fact that this rule doesn’t hold. The easiest one to get examples of 
are Presidential press conferences. A bunch of reporters are in, say, the State 
Department auditorium, and one of them is pointed to. They can ask one 
question. Then somebody else will be pointed to. This isn’t so for Johnson’s 
recent conferences because he’s been doing them informally; people wander 
around together. And there it seems to be permissable for a fellow to ask a 
second question, a clarifying question. But in general, you could take the 
Presidential press conference as one extreme: You only get one question. One 
thing you could then do is move to the other extreme: Places where one has 
an enforced right to ask further questions, a right which is not violatable by 
the person who is answering - cross examination, for example. Then you 
could try to see what questions look like under this or that control. For 
example, do they have a different shape? And what’s the position of either the 
questioner or the answerer in each situation? 

Here’s one of the problems for a questioner at a press conference. Persons 
come in with questions. They all move to be called by doing various things 

- standing up, waving their arms, getting front seats, wearing odd clothes, 
etc. One of them gets recognized. He asks a question. An answer comes out. 
If he had an opportunity to talk again, he might want to push further on what 
that answer suggests. But he can’t. Somebody else is now called on. Now their 
problem is, they may know that there’s a good question one could make up 
to follow that one. But if they follow that one, what happens to their 
question? They may have come with something of particular interest to their 
area, “What’s your policy on Idaho potatoes?’’ They want to get that out. But 
what’s just been raised is something about the defense policy, which is a 
generally important issue. Should they ask their question or the follow-up 
question? They won’t get another opportunity. Now for some questions 
they’ve brought with them, they can know that these are so topical that 
another might well raise it, and they can ‘sacrifice’ their use of it to doing 
what’s in the interest of the press conference; that is, asking a follow-up 
question. But of course the President has some control over what questions get 
asked in the first place, because he will have some pretty good idea about 


Part I 

what various persons might be interested in asking. (Kennedy apparently 
sometimes gave them the questions, or they told him what they were going 
to ask if they were recognized.) 

Another sort of problem for a questioner at a press conference is that the 
answer might be nothing more than a “Yes” or “No.” Or, for example, the 
Kennedy format, responding to a question with a quip. And that in its way 
is akin to “This is Mr Smith may I help you?” as a way of getting the other’s 
name without saying “What is your name?” and also akin to laughing when 
someone says “I’m going to kill myself’ as a way of refusing to give help 
without saying “I won’t help you.” The Kennedy quip is a way of refusing 
to answer a question without saying “I’m not going to answer that question.” 
Because after the quip, the fellow who asked the question can’t come back. 
And whether anybody else will come back to it remains to be seen. 

What might stand as a solution to this problem was something that was 
noticed when they were thinking of televising the press conferences. One of 
the matters that arose was, if they televised these things, people would ask 
very long, involved questions to get themselves time before the audience. 
Then it was realized, well they do that anyway. Now the question is, what are 
these long involved questions doing? One thing they might be doing is this: 
You try to ask a question, providing in the question for a sub-answer: 
“Would you do so-and-so or not, and if not, then . . . ?” That is, you’d want 
to provide that he could not just say “Yes” and you wouldn’t have anything, 
but “If ‘yes,’ then is such-and-such relevant, and if not, is such-and-such 
relevant?” These sorts of questions sometimes get so involved that after one 
of them the President says “Well you’ve made your point, thank you.” 
Because it looks like a speech. Now that elaborateness may not be so much 
due to the guy’s being on television as to his having to make a question under 
this constraint of not having a second chance, to ask for a clarification of the 
answer or to pose a follow-up question. You have to try, then, to build the one 
question to do that job, whereas in other situations you can just ask them piece 
by piece. 

So we can begin to see how things might look different by virtue of the 
presence or absence of the rule and how strongly that is enforced. For 
example, in cross-examination the presence of the rule is strictly enforced. 
You’re sworn in and you stay on the stand until the guys says “No more 
questions”. You can’t say “Well that’s enough, don’t bother me anymore,” 
which you can do in ordinary conversation. And, for example, in ordinary 
conversation you can produce an answer, append to it a question, and now 
take over. The roles are not set so that one is ‘a questioner’ and one is ‘an 
answerer’ from now until the end of time, or from now until the end of this 
interchange. So something as limited as this rule can have rather considerable 
and observable consequences. 

I’m going to introduce another rule. I’ll start out with some quotations. 1 

1 Cases (l)-(4) are from Sacks’ research notes. He often put big blocks of data on 
the blackboard, usually reading it out. He does not do so this time, but he refers 

Lecture 7 


(1) A: Do you have any physical problem? Any illnesses? 

B : No. Just a little bit ( ) and overweight. 

A : Overweight? 

B : Just about 20 pounds. 

A: Now when you get depressed, how is your appetite? 

B : It’s usually always good. No when I’m depressed I don’t feel like 

eating, I can’t eat. 

A: And how do you sleep? 

B : Alright, but not as much as I’d like to. 

A: That happens. Do you lose much interest in sex when you’re 


B : Well, just in everything when I’m depressed. 

A: Now, right now you’re feeling a little better. 

B : Yes. 

A: But you figure on past experience that you might become 

depressed again any time, like tomorrow. 

B : Yes. 

A ► Well, it certainly sounds as if you need some help, and there are 
many things that can be done for this. 

(2) A: Are you working? 

B : No I’m not. 

A: Your husband supports you? 

B : Yes. 

A: Well, what do you do with yourself? 

B : Oh, I have a lot of interests. I work with theater. I do, oh, little 

community theater direction and things of this order. 

A: Do you find there are times when you lose interest in it? 

B : Yes. Very d- I find there are times when I lose interest in 

everything and there have been times when I have stopped 
speaking for days. 

A : I don’t know anything about your sex life now, but are there times 

when you lose interest in sex? 

B : Yes. Completely so. 

A: - *Right. Sounds pretty clear cut as a depressive illness. 

(3) A: Hello, this is Mr Smith. 

B : How does the Emergency Psychiatric Hospital work out? 

A: How does it work out? 

B : I mean what do you do. I’m sure a telephone conversation 

wouldn’t save me. 

A: Well, it sometimes is a first step in the process, going a very long 

way in being helpful. 

to the materials in such a way that at least two of them can be located with some 
certainty, and the other two stand as decent instances of what is being talked of in the 


Part I 

B : A first what now? 

A: A first step. 

B : In a telephone conversation. Then what’s the next step? 

A : Well, we would If it’s a problem involving suicide we would invite 

you to come in for an interview, explore the problem more fully 
with you, and then see if we could recommend something helpful. 

B *-It sounds too slow. 

A: What’s that I’m having difficulty hearing. 

B : I said it sounds too slow. 

A: Too slow, we can act very rapidly when the need arises. Are you 

a person with a suicidal problem? 

B : Yes. 

A: Could you tell me something about it? 

What we find in these exchanges is that the person who is asking the questions 
seems to have first rights to perform an operation on the set of answers. You 
can call it ‘draw a conclusion.’ Socrates used the phrase ‘add them up.’ It was 
very basic to his way of doing dialectic. He would go along and then say at 
some point, “Well, let’s see where we are. Let’s add up the answers and draw 
some conclusion.” 

And it’s that right that provides for a lot of what look like strugglings in 
some conversations, where the attempt to move into the position of 
‘questioner’ seems to be quite a thing that persons try to do. (We can just note 
that in our third exchange there’s a shift in the middle, A, the answerer, 
appending to his answer a question, which B answers, and then a next 
question.) As long as one is in the position of doing the questions, then in part 
they have control of the conversation. 

Now, we do find questions followed by questions. For example, you can 
propose a request for clarification of the question. 

( 4 ) 

A: This is Mr Smith. May I help you? 

B : Yes, I heard that you help people who are on the verge of 

committing suicide or something. 

A: What was that? 

B : I said that you help people who are on the verge of committing 


A: Yes, we do, certainly. 

B *Tn what way? 

A: - ►Well, let me ask this. Are you calling about yourself? 

B : Yes, uh huh. 

A: Well, what we do is talk with them. We usually invite them to 

come into the office, try to determine what the problem is, and see 
in various ways how we can help by talking understanding the 
problem and sometimes making recommendations sometimes 
getting them in contact with a psychiatric person to help them. 

B : I see. Where is the office? 

Lecture 7 


But as compared to ceremonials - where once the slot in which the next piece 
of talk goes is filled with something else, then the ceremonial doesn’t hold the 
floor, so to speak - I don’t think that’s so for questions. For example, here’s 
one where, after two questions by someone who had been asked a question, 
the first question is answered. 

A: Hey did you talk Marcia into coming down here? 

B : Was she here? 

A: Yeah. 

B: When did she leave. 

A: About a half hour ago. 

B: Yeah, I talked her into living here with me. 

So an answerer can ask a question after a question. But what happens is, you 
get a parenthesis in the conversation. There is a question on the floor, and an 
interlude in which some question or set of questions operates, where, once that 
is answered the initial question comes back into its relevance as controlling 
what’s supposed to go next. 

As I said, as long as one is doing the questions, then in part one has control 
of the conversation. In our first four instances the person who is calling is 
doing so, presumably, for help. One difference in them is that in the latter 
two, the caller starts holding the floor with a series of questions. And by doing 
that, they have the opportunity to assess whether the proposed help will be 
something that can help them and whether they’re going to use it. They can 
say, as we get in the third instance, “It sounds too slow.” In the first two, that 
doesn’t take place. And what one has in the latter two is a situation where the 
caller is in the driver’s seat; they can hold back asking whether they should 
come in, and go about deciding that on the basis of the answers they get. 
Whereas in the first two, the recipient of the call gets the materials, on the 
basis of which they can propose that the other ought to come in for help. 

There is a lot more detail about how, in particular ways, the bringing of an 
agreement or the holding off of an agreement to come in, takes place. But the 
outline of it, by reference to this rule that persons asking questions have a 
right to make the first operation on the answers, is the first thing I wanted to 
see. You can then watch for it happening in altogether different kinds of 

Let me now focus on the fact that the one who is doing the questions has 
control of the conversation, in part. There can be a sense in which, while 
you’re asking the questions, you could not be said to be in control. I’ll give one 
of the best examples I’ve ever seen of this sort of thing. I took a course called 
Constitutional Litigation with an enormously smart man, a guy named 
Telford Taylor who was chief prosecutor at the Nuremberg Trials. The first 
day, we came into the class and there were twelve people. He went through 
the list of persons who were supposed to be there. Thirteen had signed up for 
the course. The door opens and somebody walks in, just as Taylor has finished 
going through the names. He says “Mr Jones.” The guy says “No.” Walks 


Part I 

over and sits down. Taylor says, “You’re not Mr Jones.” “No, I’m not.” So 
he’s used the correction-invitation device, asking “If you’re not Mr Jones, 
who are you?” without saying that. In return, he’s gotten “No, I’m not.” So 
he says, “Well, on the list here, Mr Jones is the only one who isn’t present, 
who is signed up.” The.fellow says, “I didn’t sign up.” Taylor says, “Well I 
ought to have the names of all the people who are taking the course.” The 
fellow says, “Well I’m not sure I’m going to take the course.” Taylor says, 
“But even anybody who just happens to be here!” “Well, maybe I’ll leave.” 

So then Telford Taylor sits and scratches his head, says “I’m sure if I go to 
the Dean’s office I’ll find a rule that says if I ask somebody for their name, 
they have to give me their name.” The fellow says, “You haven’t asked me 
for my name.” At which point the class starts wildly laughing, and Taylor 
points and says, “Now sometimes you get a witness like that!” And you do, 
of course, get just this kind of thing; the ability to extend a series of returns 
which avoid what’s being asked for but hasn’t been asked for point blank. 

Now, in contrast to that, we get something we can call an ‘answer 
constructed by reference to the project of the question.’ We’ve talked about 
this correction-invitation device. What you got there was, for example, “Is 
your husband a police officer?” to which you could answer “No” or “No, he’s 
a such-and-such.” And it’s the latter answer which we can call an answer 
constructed by reference to the project of the question. That is, what you can 
see that the question wants to find out, is something that controls how you 
answer it. Here is another instance of such an answer. 

B: I just thought you sent somebody out to talk to somebody. You 

know, Twelve Step type of thing. Because I must have a- 

A: Have you ever been in AA? 

B: I don’t drink. 

A: You don’t drink? 

B: I mean, I drink if I’m out at a party and everybody’s having a drink. 

I take one to everybody’s two, but I’ve had friends . . . 

That is, the caller has mentioned that this clinic might work like Alcoholics 
Anonymous (Twelve Steps being one of their procedures). Asked if she’s ever 
been in that organization, she gives this answer, “I don’t drink.” She could 
have said just “No.” But she understood what it was that he wanted to find 
out, and produced an answer by reference to that. 

These sorts of answers take a kind of cooperation which is not present when 
the answerer leaves the questioner ‘in control’ of the conversation - which can 
also be characterized as letting the questioner go off on as many wrong tracks 
as he pleases, where you can get a long, involved project that generates a series 
of questions, none of which turn out to have any use. 

Lecture 8 
On Measuring 

I’ll begin off with some quotations. 1 

(1) A: People have to be pretty upset to want to kill themselves. 

B: Oh I have been. Don’t think I haven’t. 

A: Have you been sleeping? 

B: No I haven’t. 

A: And how’s your appetite. 

B: Well, not too good. 

(2) A: Let me hear a little more. How’s your appetite. 

B: Not too good. 

A: Yeah. Haven’t felt like eating? 

B: ( ) other day 

A: And you haven’t been sleeping too well? 

B: Not very well. 

(3) A: How are you sleeping? 

B: Well, that’s a peculiar part, I sleep very well. 

A: And how is your appetite. 

B: My appetite is very good. That’s what I don’t understand. 

A: You sound as though you know something about the symptoms of 


I want to notice the following sort of thing. Recurrently in these 
conversations you get a question like “How are you sleeping?” “Not too 
good,” “How’s your appetite?” “Not good.” Now it looks like absolutely 
nothing is going on. And quite frankly when I thought I was essentially 
finished with analyzing this stuff, I hadn’t seen something that’s going on 
there that is utterly fantastic. Absolutely unbelievable in a way. And let me 
just try to begin to show what it is. 

I’ll start out with something I’d already worked up, which just begins to 
set this problem. And that is, there are a variety of items on which persons can 

A combination of extracts from several lectures: Fall 1964, tape 14, side 2, 
pp. 1-10, Ml, side 2, pp. 16-17, M2, side 1, pp. 1-4, 7-12, and 14-18, and 
Winter 1965, lecture (4-5), 14-15 (transcriber unknown). 

1 Only the first of these extracts is on the tape; the other two are taken from Sacks’ 
research notes. 



Part I 

monitor their own states. They either do it routinely, or, when asked, they 
have the information which permits them to engage in some consideration 
and then give an answer as to a current state or current variation from a prior 
state. The items are things like sleeping, appetite, etc. And all I’m saying, 
then, is that persons can go about monitoring how they’re eating and sleeping, 
on those questions, they know. 

Now for each of these there is a category that Members use, and that 
category is ‘normal.’ That is an extraordinarily special category. Each Member 
employs it. Questions can be asked about it from one to another. But for a 
large variety of uses of the answer to questions within which ‘normal’ might 
be an answer, and for a large variety of the monitorings which might go on 
by oneself without anybody asking you anything about it, it’s quite irrelevant 
what, for any given person, that notion ‘normal’ denotes. It’s irrelevant 
whether it’s similar to or different from the features that anybody else uses to 
decide that on some item they’re ‘normal.’ So ‘normal’ is a standardized 
category, where whatever it refers to for any given person doesn’t have to be 
specified to control its use. 

And there are variant categories which are also standardized, like ‘poor’ 
and ‘great.’ There are whole bunches of terms: Fantastic, cool, terrible, 
whatever you want. It’s essentially the same. We’ll call them ‘directional 
differences.’ Minus and plus. So we get ‘normal,’ ‘minus’ ‘plus.’ 

Variations from ‘normal’ are noticeable phenomena. They’re noticeable by 
reference to whatever it is that’s ‘normal for me.’ And it’s the fact of the 
variation which is relevant to some state being noticeable, and not what the 
normal state’s features are. That is to say, if you sleep four hours a night 
normally, that doesn’t make how much you sleep noticeable. Two hours 
might be ‘poor.’ That would make it noticeable. Six hours might be ‘poor’ for 
somebody else; that would make it noticeable for them. You don’t have to 
have an equivalence. 

It’s also to be observed, and equally crucial to the whole business, that the 
variation categories are standardized, and they’re standardized without respect 
to what they contain or what the normal contains. Any Member can employ 
the set of categories to formulate their current state; that is, they can say 
‘normal,’ ‘poor,’ or ‘great’ without reference to what it is that that stands for, 
or how what it stands for compares with what anybody else has theirs stand 
for, and they can talk about it. And if the product of some monitoring comes 
up with one of the variant states, that provides that that state is noticeable, 
and provides, then, an occasion for an account of that variant state. That is, 
it provides for an inquiry being launched as to how come it’s that. If there is 
a collection of variances, then the problem that an inquiry has to solve is their 

Now consider how extraordinarily elegant this is, as compared to a 
situation which is imaginable if one had to have a standardized content to 
these things. That is, if one had to have a standard measure such that persons 
would have to decide whether ‘my normal’ was ‘normal,’ and if one had to 
have measures for the variants, and it was required to know what the ‘normal’ 

Lecture 8 


was, for any given person to decide that ‘my poor’ was ‘poor.’ But such 
standardization is, not absolutely irrelevant, but essentially irrelevant to the 
way these measurements are actually done. 

There are very special cases, which are set up just by this thing being so 
extraordinarily simple, where these matters do come into question. That is, 
there are some physiological states which are dramatically different with 
respect to what is ‘normal,’ where that might be taken as something that 
ought to be found out about. But an enormous amount of diagnostic talk - 
with professionals, and in conversation which is not with professionals - can 
go on without much further specification than that; without having to check 
out what is ‘normal,’ what is ‘poor,’ whether that’s ‘really normal,’ whether 
that’s ‘really poor,’ etc. 

It’s that set of facts, and the set of items for which those facts hold that 
provide part of the superstructure that permits these conversations to take 
place. People don’t know each other? They don’t need to know each other and 
nonetheless all this can go on. With respect to these kinds of items we do not 
have what we ordinarily refer to as ‘ambiguity troubles.’ Asking, for example, 
“How’s your appetite?” and hearing “Not too good,” that’s sufficient. What 
it is that “Not too good” is for that person doesn’t have to be known. And, 
for example, that someone eats an awful lot normally and now what he eats 
would still be a lot for most people, doesn’t matter. 

What I’m trying to differentiate here is the following. These are clearly 
measuring categories of a sort. And one might have the idea that, given the 
way measuring is done, these are ‘rough versions’ of something that is 
mathematicalizable; that persons are talking loosely about something which 
our business would be, in studying it, to find out what the specific measures 
are. What I’m saying is, there aren’t any such measures built into the use of 
these categories. I’m pointing to the fact that here’s a medical device of a sort. 
Its power comes from its emptiness in a way. It measures directions, and it can 
work for normality apparently by reference to ‘whatever it is that normal is,’ 
and that’s enough. That there is a set of items like this is fantastic. Consider 
what a time system would look like, or what a monetary system would look 
like, built on this kind of a device. 

What we want to consider carefully, much more carefully than sociologists 
have ever considered the matter, is what kind of an object do you have for the 
purposes of counting? Here is another way I came to think about this matter. 
In about 1956, in the Journal of World Politics, Talcott Parsons wrote a review 
of C. Wright Mills’ The Power Elite. I’m not a great fan of Parsons, but there 
he made what I thought was a really basic kind of point. He said, “In this 
book Mills seems to propose that power is a zero-sum phenomenon.” And 
what that means is, if I have it, you don’t. Something adds up to a number 
that can be divided among the set of persons such that if 90 percent are X, 
then 10 percent are Y; that is, there can only be 100 percent. 

Quite obviously there are lots of things that are not zero-sum phenomena. 
For example, an economy may not be a zero-sum phenomenon; it can expand. 
If you put something in you can get more out. And if everybody has a certain 


Part I 

amount, that doesn’t mean that that’s all there is. And when you’re 
considering social phenomena, one thing you want to do is to try to find out 
what kind of a counting system you need. Is the object zero-sum or 
non-zero-sum, or something quite different? And then, how is it that persons 
go about counting this or that matter? 

Let me give a rather simple kind of consideration. Here’s a quote from a 
group therapy session of seventeen, eighteen-year-old kids. The group is made 
up of four boys and a girl. However, the girl left the group the week before. 
In this exchange, one of the boys is talking to the therapist, also male. 

A: Did Marian call or anything this morning? 

B: Why, did you expect her to call? 

A: No. I was just kind of hoping that she might be able to figure out 

some way of coming to the meetings. She did seem like she wanted 
to come back. 

B: Do you miss her? 

A: Oh, in some ways yes it was nice having the opposite sex in the 

room, ya know, having a chick in the room. 

The first thing to notice here is that MIR categorization device is being used 
to formulate the absent person, the relevant set ‘sex,’ the category ‘female.’ In 
general, it’s to be noticed that the MIR device provides one of the basic ways 
that Members go about counting all sorts of things. This can be done, since 
for any category-set there is a known set of categories, and what you do is 
examine some population - a meeting, a party, one’s employees - deciding 
whether, for this or that category-set, every category is represented. You can 
have, then, a notion of completeness and a notion of absences. You can then 
formulate the sort of thing you need to make this or that population 
complete, where you don’t have to know the people you need; you only have 
to know that they have to be ‘females,’ for example. 

In this particular case, the way this fellow finds of formulating someone’s 
absence is psychodynamically interesting. He is, for various reasons, much 
afraid of ever showing affection for anybody. And as much as he could do was 
to provide for the fact that ‘some such one’ was absent, and that was good 
enough grounds for him to miss her. But he wasn’t going to say that he likes 
her or anything like that. 

More abstractly, such a use of the MIR device is a way one can go about 
making what I think can reasonably be called ‘safe’ or ‘non-commital’ 
compliments. A category, in this case ‘female,’ is applied to someone, which 
discriminates that one from every other person in the place, so that someone 
is characterized by reference to a category that applies to nobody else. The 
relevance of that for making a safe compliment may be considered if you 
think of compliments like “It was nice having someone smart in the room” 
where, on its use, anybody else can say “Well what about me?” That is to say, 
there is a whole range of categories which also can apply to any other person 
in the room, and thereby if one is singled out for notice on that item, the 

Lecture 8 


others may be implicitly derogated. Another thing is, just to mention it, that 
on the category-set used to provide for the way that she’s located, it’s not 
simply the case that she’s a member of a category that nobody else is a 
member of, but that everybody else is a member of the same category by 
reference to the one that’s been used to locate her. 

Now, there are whole series of ways in which we can talk about Members’ 
counting, measuring, adding up, etc.; that is, we can go about mathematizing 
social events. Here’s an interesting way in which this matter comes up. A 
woman is talking to her husband; their son had a birthday that day. 

A: I said did Grandma bring you anything for your birthday, I forgot all 

about it. And Nicki said uh 

B: Yeah I did too this morning. 

A: Forgot all about it. 

B: Oh boy. Well. 

A: I said, Nicki I said Mama has so much on my mind I forgot all about 


What we can see here is how some set of things can exclude some other thing. 
The issue wouldn’t be that one had 18 things on one’s mind and therefore by 
some logical operation the 19th could not be held in mind. We need to ask 
what it is that stands as excluding something, when and how any given set of 
things can provide for the exclusion of another thing, or can substitute for 
something - all of which are mathematical operations. 

Then there is a very nice issue, that of ‘erasability. ’ Some things are in a 
sense erasable; for example, one can be forgiven. But in another sense it can 
be seen to remain there, on the sheet as it were, such that if that thing happens 
again, someone can say “I already forgave you once.” Other things don’t 
seem to have even that sense of erasability. Take, for example, suicide 
attempts. For many diseases, let’s imagine that the record of your life is a 
hospital log with a series of entries. Now let’s say there’s an entry: “Had 
pneumonia on such-and-such a date. Cured.” That’s that. If you get 
pneumonia ten years from now nobody says “Gee, how come they didn’t cure 
him of pneumonia?” You can get it again and this prior one had nothing to 
do with it. 

So in that sense, pneumonia and many diseases are erasable. The cure 
removes it from the record. And the next occurrence is not seen as explainable 
by the earlier one. But suicide is not that way. If you make an attempt at some 
point in your life, then the fact that, for example, you went through therapy, 
or you were seven years old and now you’ve grown up, does not remove it 
from the record. That thing remains alive. So that if you kill yourself 40 years 
later, persons will refer back: “See? He made a series of attempts.” It’s not an 
erasable matter. 

One relevance of suicide not being erasable is the following. Let’s say you’re 
now considering whether to go into therapy, as compared to killing yourself, 
where this would be your first suicide attempt. If you just do it, and do it in 


Part 1 

a way that might not look like suicide, your family won’t have stigma, can 
collect insurance, and all the rest. You’re in a situation quite different than if 
you go for help. If the help fails, thereafter you can’t kill yourself in the same 
way. And people know that. And that provides a real trouble about getting 
people into therapy. Because if they come in under suicidalness, just by going 
in they have cut off what they see as an opportunity. Because what happens 
is, if you die in some way which might not look like suicide, the fact that you 
had that record provides that people ought to look at this death by reference 
to that record, and they may decide “Suicide.’’ 

If you keep suicide attempts off your record, then if you do kill yourself, 
you may have something important going for you. And that is, the 
procedures employed for deciding that somebody committed suicide. There 
are kinds of events which are well recognized to occur. For example, 
somebody can sit down and drink themselves to death. However, it’s a very 
hard thing to do, and if pills are available, people, it’s said, just don’t do it 
that way. 

Now if you had a statistical operation for deciding did so-and-so kill 
himself, you’d take 100 cases and you’d have a notion of the probability of 
events occurring, so that in 5 percent of the cases where persons died of 
alcohol they may have killed themselves. You could then bunch those cases 
and you’d get a distribution at the end, where 95 percent of deaths by alcohol 
are accidental and 5 percent are suicide. You could pick the five cases that 
looked closest to being suicide and propose them as probable suicides. It 
doesn’t happen that way. Cases get decided one by one. And each time a case 
comes up, the fact that there’s a 95 percent chance that any death by alcohol 
is not suicide, is used to decide, “No.” And you get the statistical outcome 
which is that they never occur. 

My own feeling about such matters is that a range of decisions are made in 
terms of ‘odd events’ versus ‘normal events.’ And odd events, by and large, 
are just not added together. So that if one has a notion that some X is a 
normal event, then the fact that occasionally or two or three times in a row 
something else happens, that doesn’t provide for a shift. One doesn’t now say 
“Well, maybe X isn’t the normal event.” But, in part perhaps by way of the 
fact that what is normal gets incorporated into things like proverbs and 
becomes very stable, odd events are just sloughed off. They don’t get 
incorporated. The fact that it’s odd is enough to mean that one doesn’t have 
to consider it on this particular occasion. What you get is, “Those things 
happen, sure, but ...” 

One thing about odd events, then, is that they’re very hard to report. This 
fact can occasion the relevance of a category that this society has, called 
‘believing.’ So, for example, there are classes of events which are very closely 
tied to activities that someone ought to do. If the event occurs, the activity 
ought to be done. Now for some events that are so tied to an activity, those 
who ought to do that activity may not be in a position to observe the event, 
but they have to deal with a report. Somebody says the event happened. Fire. 
Wolf. There’s a kind of gap, in the sense that one can’t wait to see whether 

Lecture 8 


the thing that stands as a condition for some action has happened, but one has 
to act on the basis of a report. And of course the Cry Wolf fable is the fable 
about this problem. 

What’s to be done when if something happens some action has to be done, 
and nonetheless the persons who have to do that action can’t see to decide? 
And that’s of course where ‘believing’ fits as a category. ‘Believing’ can 
operate on such a basis as the following. A notion of credit is built up about 
somebody’s reporting of events. And his credit is used to decide whether the 
action ought to be done. It might be that somebody has a specific task, 
reporting event A, for which some action should be taken. And if they report 
it wrongly a series of times, they’re no longer creditable on event A. It can also 
be the case that ‘believing’ does not become event-specific, but that for some 
person, whenever they say something happened for which some action should 
be taken, whatever the event and its relevant action, they just don’t have any 
more credit. It may be they have no more credit across whole sets of classes 
of events, or it may be that while they’re not believable in general, for some 
events you’d figure this is not the kind of thing anybody lies about. And if one 
was interested in mathematizing social events, one would want to find out 
what the classes of events are, when they overlap, etc. For example, which 
kind of event is used to decide that persons are unreliable on another kind of 
event, and which are not. 

Now one of the things about ‘believing’ is, if you can propose that a person 
has no credit, then it seems to be quite adequate to use the fact that you didn’t 
believe them, even though the event they reported did occur. In my suicide 
cases there are situations where a person has killed himself, investigators ask 
a friend or relative or neighbor, “Did he ever threaten?’’ They say, 
“Threatened all the time but never did it, so we never took him seriously.” 
And that’s treated as quite an adequate account by the persons involved, and 
apparently by those they tell it to. 

Well, that’s fine, if what is at issue is building an adequate account. But 
suppose you have the following problem - and it’s a live problem. You’re 
setting up a missile receptive system, where dots appear on a radar screen. 
And the people who monitor the radar screen have to tell somebody to act. 
Now ‘believing’ is the resource the society has. But the one who has to act can 
know that the screens can show dots when there’s nothing there, or can show 
dots and there’s something there but it isn’t missiles. And if you believe and 
act upon the reported dots, you’re in plenty of trouble. And if there are dots 
and they are missiles, and you haven’t believed the reports, then you’re also 
in plenty of trouble. But that ’s the category that people seem to be stuck with. 
And of course in this case an account will obviously not be adequate in the 
sense that, whether there were no missiles and you acted, or there were 
missiles and you didn’t act, nobody is going to be around to say, “Well, so 
you did your best.” 

I’ve heard talks by persons who are in the business of, I think they call it 
‘human systems’ for things like the Air Force, and as I understand it, this is 
an actual problem that they don’t know quite how to deal with. It seems 


Part 1 

they’ve tried some experiments designed to select persons whose reports 
would be reliable. And that would be decided by how persons behaved in the 
experiment. Now that would seem to be fairly reasonable. Except that these 
guys know it’s an experiment, and now they’re going about considering what 
happens if they behave too conservatively, will they ever be chosen? And if 
they’re not conservative enough, will they ever be chosen? And that of course 
affected the outcomes of the experiments. 

Now, for some sorts of organizations there seems to be a system for 
deciding reliability that has to do, not with behavior but with category 
membership. One day I was going through some police files on suicide and 
I came across a series of notes to the police involving the discovery of a body 
in a college dormitory. A fellow was found hanging in a closet. Here’s the 
sequence of notes. 

The first note says, “I am Mrs so-and-so. I am a housemaid in the 
such-and-such dormitory. At 10:08 in the morning I was cleaning out room 
472. I had swept around the room. The closet door was slightly open. I 
opened it a little more, and there was somebody hanging there. I immediately 
stopped. I went down and found my supervisor, Mrs so-and-so.” 

Next note. “I am Mrs so-and-so, supervisor of such-and-such floor in 
such-and-such dormitory. Mrs so-and-so came to me at such-and-such time 
and said that there was a body in the closet of room 472. 1 went to room 472, 
looked in the closet, and there was a body there. I called the head of the 

And then a note from the head of the dormitory. “At 10: 14 I was called 
by Mrs so-and-so, the supervisor for such-and-such floor, who said that there 
was a body in the closet in room 472. I took the elevator up, went into the 
room, looked in the closet, there was a body there, which I then cut down. I 
then called the police.” 

You could figure my God if you want to talk about organizations and 
hierarchies, isn’t that something? That on such a matter that’s the way it 
operates. Well, but maybe it’s occasional. Maybe it just happened that time. 
No so. Recurrently when a body is found in an insitution, you get that 
sequence. I routinely clip the newspapers, and when you look at things like 
murders, suicides, etc., you find things like this. There was a recent murder 
in New York, and the report says, “The body was discovered at 5:24 a.m. by 
the janitor who woke up the superintendant who then went and checked and 
then called the police.” Some categories in a hierarchical organization do not 
have enough rights to say that there’s a body somewhere without some other 
category checking to see that there is indeed a body somewhere. Nobody 
touches the body until somebody who is ‘responsible’ comes. And they’re the 
ones who bring it into contact with the outside world. 

Now, such a sequence is not reserved to those sorts of organizations. And 
it can sometimes operate in extraordinarily tragic ways. There is, for example, 
the possibility that if the fellow who hanged himself in the college dormitory 
was alive when he was first found, he might have been saved. But the first 
time I ever thought about this matter, the following happened. I was in a 

Lecture 8 


hospital waiting room. They came charging in with this dead eight-month- 
old baby which they were trying to revive, the young mother, about 18, and 
her parents. What had happened was, so the mother thought, the baby had 
suffocated by a pillow over its face. Actually the baby had some disease, but 
the mother thought it had suffocated. The mother and baby lived about five 
blocks from the hospital. When she discovered the baby was not breathing - 
she happened to walk into the bedroom and noticed this - she goes to the 
phone and calls up her parents who live 45 minutes away, sits down on the 
front steps, waits until her parents come, and they call the police. And the 
police then call the ambulance. 

These built in sequences! I’ve been interested in the possibility of 
orderliness in what people do for a long time, and not known that they could 
be working in that kind of order. Now, the way I work has been called 
‘microscopic’ with, then, the usual sociology as ‘macroscopic.’ And it’s not a 
bad distinction. But then it’s proposed that social events are not closely 
enough ordered so that we can get results at the ‘microscopic’ level of 
investigation. I take it that we just don’t know whether or not that is so. 
Certainly there has been an argument, and certainly the statistical position has 
won out. Durkheim posed the matter - which is the basis for the statistical 
approach to sociology - that if you take the statistical figures on suicides, you 
find quite an order. And you can study those. And construct theories. Then 
he says if, however, you deal with such things as the accounts that accompany 
each suicide, you don’t find order at all. They’re hastily made up, by low 
grade officials, etc. But he did not in fact attempt to deal with the accounts. 
And, in fact, his arguments as to why you shouldn’t do non-statistical work 
were statistical arguments. And it may well be, for example, that the accounts 
of suicides are closely ordered phenomena. 

Another objection to the way I work is that it seems to be enormously 
laborious, and sociologists are not given to doing things slowly. But one 
reason I’m operating as I do is, I take it that the big problem is not that we 
know that social events are not closely ordered, but that we wouldn’t know 
how to describe them to see whether they’re closely ordered or not. I want to 
see if such work can be done in the first place. Then we can repose the issue 
of where are social events closely ordered and where are they not closely 
ordered. But it’s then an empirical issue that has to be discovered. It couldn’t 
be solved by an argument. And we’ll see. 

Lecture 9 
“I Am Nothing” 

B : I’m a grown woman an attractive woman I have a 

^real nice date 

A: Do you have any (at same time as) 

B: good looking guy for a date tonight and and I somehow I’m feeling 
that I’m nothing (smiling sigh) 

A: Uh huh 

B: And I know nobody’s a nothing But I am. It’s like everybody 

else is somebody or something and somewhere along the line I muffed 

Suicidal persons recurrently say about themselves such things as “I’m 
nothing” or “I’ve got nothing.” The question I’m going to address is, can we 
describe how they’re able, properly, to so speak? Is there some way we can go 
about constructing the procedure whereby ‘nothing’ is a possible product? 

We might first ask what would the relevance of such a procedure be for the 
persons involved? Maybe for suicidal persons it is said of them, and they know 
it is said of them, that they and the project are irrational. If there were 
procedures available so that they could report how they arrived at the 
conclusion “I am nothing” or “I have nothing,” and the person they were 
talking to could go through the procedure again and find that their conclusion 
was correct, then that might provide that at least this part of their project is 
logical and not irrational. 

Or, for example, if to arrive at this conclusion one uses materials that are 
properly to be considered in producing an activity, and handles them 
properly, then another claim that is raised against persons who are suicidal can 
be undercut. The claim has been put by Menninger who says about neurotic 
or psychotic persons in general and about suicidal persons in particular, that 
they lack a loyalty to reality. (Which suggests that ‘reality’ is, for this society, 
a special category. Some scientists and philosophers might say that whatever 
is, is ‘real,’ and the category ‘reality’ encompasses whatever happens.) A 
person might use a procedure which is otherwise properly used to make 
assessments, to arrive at the conclusion “I am nothing” as a warrant for 
suicide, so as to show that they are committed to what the society holds is 
important or sacred, and that it’s out of just this commitment that the project 

A combination of Fall 1964, tape 7, side 2, tape 8, sides 1 and 2, and Winter 
1965, lecture (6) with a brief extract from Winter 1965, lecture (4-5), pp. 1-3. 
Transcriber unknown on all these. 


Lecture 9 


arises: “I’m not disloyal, I’m forced to conclude that I’m nothing.” And 
persons recurrently do propose that the suicide possibility is becoming very 
logical: “I tried to reason it out and that’s the way it looks.” 

If we’re going to construct these procedures whereby “I am nothing” is 
arrived at properly and reproducibly (i.e., others could go through the 
procedure again and find that the conclusion was correct), a first thing we 
might ask is, what are the objects by which Members in general make 
assessments of their lives? We’d find that those objects are some subset of the 
society’s recognized values; in particular, those which have a sanctionable way 
of being counted. 

Of those there is a special class of values, which are what I call 
‘cumulative.’ These have two features; (1) If you can properly have them at 
some stage of your life, then you ought to have them at any time thereafter. 
(2) You can lose them, so that having them at one stage of your life is no 
guarantee that you will have them thereafter. And the way they’re counted 
provides that the fact that you had them once is quite irrelevant if you don’t 
have them at the time you’re doing the count. The fact that you had them 
once doesn’t count. 

That differentiates between things like money and children, and things like 
‘kicks.’ For example, the cumulativity of the value ‘having children’ is 
recognized in the proverb ‘Children should bury their fathers.’ If you had 
children at some point, you ought still to have them; if they are now dead or 
if an irrevocable break has occurred, they don’t count. Money is counted in 
the same way. When persons assess their circumstances they don’t feel that 
they may add up how much money they earned over their lifetime, but 
consider how much they have now. If they do talk about how much they’ve 
earned in the past - if they don’t have any now - then they’ll talk about it 
by reference to a disparity. 

Aside from values like these, for which one can say “I have none,” there 
are other sorts of things which are cumulative, but on which you get a yes-no 
alternative; for example, being married. For such things, ‘no’ is apparently 
equivalent to ‘nothing,’ and ‘yes’ at least to ‘something.’ That is, just because 
a thing does not have the countable property of money does not make it 
unassessable in the sense we’re examining. 

Kicks are different. We can sanctionably say, “Well I had a lot of kicks.” 
The fact that we’re having none now, or few, doesn’t seem relevant. There 
may be other things of this sort. For example, if you do something that counts 
in the society - an invention, a discovery, a contribution - then that can follow 
you. At any future time, should a person who’s done such a thing try to make 
the conclusion “I’m nothing,” it can be countered with “You’ve done this.” 

Now, some of the assessable values are so structured that they are only 
relevant at a certain stage of life. For example, a caller to this emergency 
psychiatric hospital talks about himself for awhile and then says “I feel I’m 
a bum. I’m just nothing.” And the fellow answers, “No, you’re not a bum, 
you’re not old enough to be a bum.” That is, that’s one thing you’re not 
eligible to be at this stage, so you can’t use that to count yourself as ‘nothing.’ 


Part I 

So there’s a notion of a stage in life in which you’re entitled to say whether or 
not you have nothing on this or that value. When persons 2 5 years old say in 
assessing themselves that they’re unmarried, they’re told, “No, you can’t say 
that yet.’’ That’s not anything that counts as ‘nothing’ at this point. These 
things are standardized; it’s a matter of certain formal properties, that your 
age has to be X before Y counts as ‘nothing.’ This is without regard to the fact 
that a child, for example, could be said to be ‘unemployed’ or ‘unmarried.’ 
It may be true, but it’s not relevant in this society. 

But to get a sense of this ‘stage of life’ feature as a formal one, we can note 
that not all that long ago, a quite young child could be relevantly 
characterized as ‘unemployed;’ say, before the child labor laws. And, for 
example, there are old neighborhoods in New York which are partly 
populated by descendants of Spanish Jews. Among them the proper age to 
get married is about 15 or 16. Girls who want to graduate high school - not 
to mention college - find themselves under tremendous pressure from their 
families, and themselves. That is, the girls themselves talk ambivalently about 
behaving quite progressively when they’re 19 years old and aren’t married — 
or as being failures. And there are Jewish guys in the neighborhood, but not 
Spanish Jews, who date these girls and find it tremendously troublesome 
because they are too well received by the families. The fact that you’re 17 or 
18 makes you perfectly eligible to get married to their daughters, and if 
you’re taking them out, you must be prepared for that. 

Related to the stage of life feature is that of ‘prospects.’ A person who 
engages in the assessment procedure I’m describing, and arrives at “I am 
nothing,” may then find somebody saying, “Well, that’s where you are now 
but what about your prospects?” Or, as we find consistently for any relevant 
question, this question could be expectable without ever being asked; that is, 
it can be independently dealt with by the person doing the assessment. So how 
do they deal with the matter of ‘prospects’? Here’s some further talk by the 
woman I quoted at the start of this conversation. 

B: And I do think what man wants a neurotic childless forty year old 

woman No man. 

A: Well I’m not 

B: I know that 

A: sure about that 

B: I know that outside of to hit the sack with 

A: Uh huh 

B: That’s not what I want 

A: Uh huh 

There seems to be a notion of ‘entitlement,’ whereby one examines some 
problematic outcome by reference to the Members’ theory of how that 
outcome is arrived at. When I talked about the MIR device, I mentioned that 
arrangements of its classess provide pathways which persons can use to talk 
about what they’re entitled to expect; given where they are now, where they 

Lecture 9 


could come out. Such a use of that device can involve formulating one’s 
present state in such a way that for the collection of categories one is presently 
in, one has no legitimate expectations of ‘getting there. ’ One can then propose 
that one has ‘no prospects’ on that item. 

In this particular case, what is being dealt with could be called a theory of 
how it is that men go about selecting a mate, and who is eligible to be 
selected. And there are some very specially interesting things going on in this 
statement, “. . . what man wants a neurotic, childless, 40-year-old woman? 
No man.” What has to be seen about that formulation, “No man,” is that 
it’s to be understood as ‘no man Member.’ That is, we’re always talking about 
classes and class members. That’s important, because you can propose that 
any actual man who would go about selecting you is not ‘a man.’ Or, if he 
is ‘a man,’ if he claims to be selecting you, that’s not what he’s up to; that is, 
he’s not up to selecting a wife, but, as she says here, he just wants someone 
“to hit the sack with.” So this business of finding oneself unentitled is not 
simply a way of assessing some possible future, but a way of dealing with any 
currently encountered person to find either that they’re not eligible, or they’re 
just using a line. 

In general, the notion that you can’t get there from here has to do with the 
ways in which persons arrange membership classes to do things; to find pairs, 
to get appointed to jobs, to become successes, etc. Such arrangements are 
taken as ways of assessing whether it’s possible that you could do that, and as 
I said, it works not simply to argue that you haven’t any prospects, but also 
as a way of deciding what it is that’s happening to you. 

Further, if it actually happens that somebody marries you, or you get a job 
or a promotion, there is a notion of ineligibility that can get used even though 
it seems that you’re eligible by virtue of having been selected. And that’s the 
notion of ‘phoney.’ “I’m a phoney” focuses on just that sort of thing. The 
most exquisite statement I’ve ever seen of that sort was written by a woman 
in the course of a psychosis which has been named ‘depersonalization.’ She 
says, “. . . the feeling itself is one of unworthiness, in the way that a 
counterfeit bill might feel when being examined by a banker with a good 
understanding and appreciation of real currency.” 1 

We might go about describing this in the following way. For any MIR 
device category, there’s a set of inferences attached to it, which are ‘common 
knowledge.’ Some person is a nominal member of a category, but feels that 
the set of inferences that are properly made about that category are not 
properly made about them. They can count down along that list of inferences 
and find “I’m not this, I’m not that,” by reference of course to a rule of 
relevance as to what one is or is not. That is, one is not just saying “I’m not 
the sheik of Arabia” or something like that, but: “I’m not one of the things 
that, given my categorial membership, I ought to be.” And that again 
provides for the difficulties such persons have in getting into a relationship 

1 In the lecture, Sacks does a rough paraphrase. The exact quote comes from the 
research notes, and is in Psychiatry, 23 (1960). 


Part I 

because they have the notion that what will happen over the course of any 
relationship they get into will involve the discovery by the other person that 
each of those inferable facts do not hold with respect to them. 

Let me just mention a related category, that of ‘imitation.’ I’m not using 
the term with respect to actors in stage performances, but where persons 
observe some activity and say, “He’s imitating.” I came across an extraordi- 
narily interesting use of this category in some of the older ethnographies, 
dealing with the situation of Negroes in the pre- and post-Civil War periods 
in the South. Again and again I found references to the activities of Negroes 
as ‘imitating whites.’ And they were characterized as being ‘marvelous 
imitators.’ Such reports are very similar to the way the behavior of children 
is characterized. And in that sense, then, it’s being said of Negroes that they’re 

Now, we want to ask what does ‘imitation’ consist of, procedurally? How 
is it that some behavior is seen as ‘imitation? One of the central things that 
seems to be involved is this: When one normally deals with the activities of 
a Member, apparently one takes it that they have some right to do some class 
of activities, and that when one engages in making out what they’re doing, 
one takes it that what one sees them doing is what they are doing. ‘Imitation’ 
seems to involve a way of characterizing some action which somebody does 
when they are unentitled to do that class of action. And if you watch the way 
the Negro slaves got talked about, or the way the emerging Negro is talked 
about, you can see how marvelous a category ‘imitation’ is, because it turns 
out that everything whites can do Negroes can imitate, but they can’t do any 
of these things that whites can do. 

So ‘imitation’ becomes a category which involves the construction of a 
parallel set of knowledge for those unentitled Members, where it doesn’t 
happen that as they do something one finds that there is ‘the doing,’ but as 
they do something one finds that they’re able to imitate. One doesn’t see that 
thing which would, by reference to the category ‘knowledge and capacity,’ be 
taking place; that is, one sees a child ‘behaving like an adult;’ following adult 
rules of etiquette, being able to produce sentences like an adult, “talking like 
a big girl,” etc. 

It’s noticeable in relation to this, that if the capacities of some persons are 
treated in this way, then one finds that certain sorts of accounts that can be 
applied to Members in general cannot be applied to them. So one finds that 
they can’t be found to be ‘responsible’ for what they do, in a non-trivial sense. 
If you watch, let’s say, the way that children’s suicides are described, you’ll see 
that it’s not enough to say that the child was depressed, the child got a gun 
and shot itself or whatever, but there’s an added item: How it is that the child 
was free at that time to do it; for example, their mother went out of the room 
and then the child took the asprins. That is, part of the causal account is some 
competent person’s actions which permitted this thing to take place. 

I take it that ‘imitation’ is one of the basic categories one wants to focus on 
when one talks about the phenomenon of ideology. Because it makes 
noticeable that there are a whole range of things that these persons obviously 

Lecture 9 


can do, which are by and large not seen as things they can do in the sense of 
things that a Member can do, and the addition of capacities is treated as more 
things that can be imitated. And that’s an extremely interesting kind of 
blindness, if you want to put it that way. It shows you the power of this 
procedure because it’s apparently a perfectly consistent and reasonable way to 
talk, and the materials are thus never shifted over to be seen as, “They can do 
those things.” In rather more abstract terms, we can come to see a way in 
which such categories as ‘imitation’ and ‘phoney’ provide us with something 
very central, in that they serve as boundary categories around the term 

Lecture 10 
Accountable Actions 

I’ll start out with several quotations. 1 

(1) A: Hello. This is Mr Smith. 

B: Hello. I was referred to your office by Mr Jones from the 

Conciliation Court, and I felt perhaps someone there would make 
an appointment for somebody there to talk to me. I don’t know 
what I want to say to you except I’m confused and the trouble - 
Ask the questions. I can answer them . . . 

(2) A: Hello Mr Brown, this is Mr Smith. 

B: I was told to call down from the Conciliation Court and speak to 

someone. He gave me a card introducing myself to whoever I was 
supposed to speak to, and we had been recommended to Family 
Service I believe it was, and they said they couldn’t do anything 
until I had talked to someone at the Emergency Psychiatric Clinic 
for psychiatric examination or something. 

A: Why did they refer you to us? 

B: Because I had tried to commit suicide. 

(3) A 








Mr Green? 

Yes sir, 

This is Mr Smith may I help you. 

Yes sir, it was suggested by Miss Geno that I call you. 

Geno. With the Epilepsy League. 

Oh yes. 

We have a problem with my daughter. 

(4) B: Hello. 

A: Mrs Gray? 

B: Yes. 

A: This is Mr Smith of the Emergency Psychiatric Clinic. 

A combination of several lectures: Fall 1964, tape 6, side 2, pp. 5-17, tape 7, side 
1, p. 1, M6, side 1, part 2, and Spring 1965, lecture 3, pp. 3-10 (transcriber 

1 Only the third fragment is actually given on the recording. The others are taken 
from Sacks’ research notes. 


Lecture 10 


B : Yes. 

A: I spoke to your daughter who was quite concerned about you and 

I wanted to talk with you and see if we could help in some way. 

(5) A: Hello Ronald 

B: Yeah. 

A: My name is Smith. 

B: Uh huh- 

A: I’m a psychiatric social worker with the Emergency Psychiatric 


B: Yeah. 

A: Your father called us this morning and was very concerned for you. 

(6) A: This is Mr Smith and I’m one of the social workers at the 

Emergency Psychiatric Clinic. 

B: Oh yeah. I would like to have an appointment if I can. I’m kind 

of a depressed personality, and I’ve been under psychiatric care 
for- I had been at one time, and I thought possibly some 
therapy might kind of help snap me out of it. 

(7) A: Hello this is Mr Smith 

B: Say, my husband is suicidal and, I mean, he’s attempted it about 

a half a dozen times . . . 

I’ll begin off by noticing that these phone calls seem to be ‘accountable 
actions.’ By that I mean that by and large on the first opportunity to talk after 
greetings, the person who’s called gives an account of how they happened to 
make the call. We can also notice that they are ‘symmetrically accountable.’ 
That is, if somebody calls the agency, the somebody calling gives an account, 
as in the first three quotations. If the agency calls somebody, the agency gives 
an account, as in the fourth and fifth. 

Now, as I mentioned once it seems to be a fact about invitations that at 
some point in some relationship, invitations are the proper way that activities 
get started, whereas a measure of intimacy between the persons is that they 
don’t use invitations. 2 The same seems to hold for accounts of phone calls. 
Persons who call up someone they are not intimate with, often construct 
accounts of how that came about. In fact, the way I began to focus on these 
things in the first place was coming across something I thought was 
extraordinarily strange, but I guess it isn’t. It’s a phone call between a man 
and woman who don’t know each other particularly well. At some point he 

A: Oh I was in the bio med library and I had big intentions of working 

all day and they flicked the lights and kicked me out. Well that’s just- 

2 See lecture 2, pp. 18-19. 


Part 1 

I was just gonna call and see if you and your husband would like to 
come over. 

He constructs the operation of the university as providing for how it is that 
he happens to come to make this phone call. There he was, he was going to 
study, then he gets kicked out, he’s left with nothing to do, and now, by 
virtue of that, he comes to make the call. It seemed a little elaborate to me, 
but persons do that. And at a little later point in intimacy it seems that they 
don’t. They just call up and say “Hi, how are you.” And if you ask “Why 
did you call?” they say “No reason, just felt like calling.” 

In the calls to the emergency psychiatric clinic, some of the accounts that 
persons offer involve another organization proposing that they make the call: 
“I was referred to your office by Mr Jones from the Conciliation Court,” “It 
was suggested by Miss Geno of the Epilepsy League that I call you,” etc., etc. 
Some of the accounts have a different character, simply announcing the 
trouble: “I’m kind of a depressed personality,” or “My husband is suicidal.” 

Now, for the purpose of getting a hearing, and perhaps for the purpose of 
getting help, there is a sense in which the two sorts of accounts are 
substitutable. You can use one or the other, and either of them can work. But 
I want to notice that the use of each of these has a different source; that is to 
say, different search procedures would have generated the organizational 
reference and the announcement of a trouble. 

To call an organization and propose that the call takes place by virtue of the 
operation of some other organization, involves informing those you’re now 
calling that this action is not the first step in the search. That’s non-trivial 
because persons tend to start a search for help with that place about which 
they have the most rights to have an expectation that they’ll get help. That’s 
very unfortunate in a way, because if they don’t get it there, they may apply 
a formula to that fact: “If not there, where I had most rights to expect help, 
then where can I expect help from?” That is, they treat the sequence of calls 
as informing them about the likelihood that any next call will be a success. 
And they may, then, give up. I have mentioned a few calls in which that’s 
proposed. 3 In one, for example, a woman says, “I’m a Catholic and you 
know, I wonder about Christianity. I just called three Catholic sources. None 
of them offered anything.” In another, a fellow calls up and says, “I’m very 
ashamed to be a Jew,” and goes through a report of the difficulties he had 
when he called a Jewish agency. He says, “Now there’s nothing you can do. 
I’m going to kill myself. This treatment I got was the last straw. I only want 
it to be known why, and that’s the reason I’m calling. You should tell them 
they ought to treat people better than they do.” And I talk about the way in 
which persons take it that when they call an organization the treatment they 
get is representative. As this Jewish fellow says, “It’s not that they did it to 
me, they don’t know me, and I don’t feel personally affronted. But that they 
treat people that way!” And at least some organizations have a solution to such 

3 See lecture 6, pp. 43-4. 

Lecture 10 


a problem. They can come off without having refused help by proposing that 
they are giving help, where the help they are giving you is telling you to call 
this other agency which is specially capable of doing it. 

Let’s turn now to the other sort of account; those in which persons propose 
straight out that they need help. With respect to how that would have come 
about, a first thing that seems to be worth considering is how asking for help 
is regulated. In an earlier discussion I drew a distinction between two classes 
of ‘others,’ and said something to the effect that one class were persons with 
respect to whom there was such a bond of obligation that you could directly 
say to them “I need help,” for example, ‘family.’' 4 We can talk of the two 
classes as ‘insiders’ and ‘outsiders.’ The issues involved here are extraordi- 
narily interesting, but to consider them fully I have to introduce a series of 
other concepts. I will deliver a long discussion on the matter eventually; for 
now, I’ll just sketch it. 5 

There are strong feelings - and strong maxims which provide those feelings 
- that you don’t turn to “a stranger” for help with personal troubles. And you 
certainly don’t as a first act. There’s a proper sequence. And the first place you 
turn is to your family, or to some insider with respect to yourself. When 
somebody calls up an agency and says “I need help” - which is the proper 
thing to say when you’re talking to an insider - a signal ought to go up that 
something is going on here. What can be present is, they’ve turned to an 
insider and gotten no help, or for some reason they find that they’re not able 
to turn to an insider. And in this regard, something which might stand as just 
another way that callers to this agency start off, may be seen as an alternative 
to ‘‘I need help.” Things like this: 

A: Hello. This is Mr Smith 

B: How do you spell your name? 

A: S-m-i-t-h 

B: I would like to have some information about your Emergency 

Psychiatric Clinic. 

A: Hello. This is Mr Smith 

B: How does this Emergency Psychiatric Clinic work out? 

A: How does it work out? 

B: I mean what do you do. 

A: This is Mr Smith. May I help you? 

B : Yes, I heard that you help people who are on the verge of committing 

suicide or something. 

4 See lecture 2, pp. 14-15. 

5 The several lectures in which these matters are discussed are not included in this 
volume: Fall 1964, tape 10, sides 1 and 2, tape 11, sides 1 and 2, and tape 12, side 
1. For a formal version see H. Sacks, ‘The search for help: no one to turn to’ In E. 
Schneidman (ed.), Essays in Self-destruction (New York: Science House, 1967), 


Part I 

A: What was that? 

B: I said that you help people who are on the verge of committing 


A: Yes, we do, certainly. 

B: In what way? 

That is, we get callers starting off with ‘requests for information’ which 
involve, then, that over the course of the conversation help will be offered 
without being ‘asked for.’ Because asking for help is something that one 
doesn’t do with respect to strangers. So, “I need help” and “How does this 
organization work?” seem to be alternatives in the sense that persons who use 
them have perhaps come to do this call by the same paths. 

Now, if a person does turn to an insider and does get help, it’s notable that 
the form of help may be that the insider calls the agency for them - thus 
setting up the sort of thing we see in our fourth and fifth extracts. Or the 
insider may refer them to the agency. And calls made on that basis will take 
a very similar format to the calls where the person says “Such-and-such 
organization told me to call.” For example: 

A: This is Mr Smith. May I help you? 

B: Well, I don’t know. My brother suggested that I call you. 

A: I see, well he must have had some reason for making the suggestion. 

Has there been some personal problem or difficulty that you’re 

B : Yes, I just lost my wife and I feel awfully depressed. 

And there’s another which goes, “A friend of mine told me to call.” “Why?” 
Caller says something, and then the receiver says, “She certainly did the right 

I want to stop here and show a way to analyze classes of statements for sorts 
of things we can find in them. I had these exchanges, where a person calls and 
says “My brother suggested that I call you” and “A friend of mine told me 
to call.” I thought gee, that’s curious. And I tried to think of a paradigm line, 
of which ‘X told me to call’ is an instance. It looked like the kind of line about 
which a psychiatrist would say, “This is a dependent person.” And then I 
tried to see what it was that made it look like a matter of ‘dependence.’ I 
thought of the following. A kid comes into a grocery store and says “My 
mother told me to buy a dozen eggs.” It looks exactly alike. It looks very 
much alike, anyway. And that’s one of the major prototypic types of 
childrens’ accounts; naming some adult who told them to do some activity. 

Parenthetically, we can note that the fact that such an account by a child 
is sufficient and they learn that this is sufficient very rapidly, causes rather nice 
trouble. They can go from one to another among the set of adults who can 
warrant some activity, to find one that will say “Okay.” And I have some 
stuff in which children were watched over a period of time, and we find that 
they do systematically go from one to the other without telling the one that 

Lecture 10 


they asked the other, or what response they got, so as to provide that one of 
them will say “Okay.” Then they go back and say, “He said okay.” And of 
course that’s known to generate fights in the family, etc. 

So we get this prototype, ‘an adult said to do it,’ and something much like 
it done by callers to these agencies. And what it does, of course, is to provide 
that some person other than the caller - or in more general terms, the doer of 
some action - is the competent individual in the case. Now, apart from the 
fact that a person is proposing that they’re not a competent actor where as 
adults they might be expected to be competent with respect to some item, the 
importance of this is the following sort of thing. If one can formulate what a 
child’s set of resources are at some point, then one may be able to examine 
those behaviors of adults who are said to be ‘infantile,’ to find just what it is 
about them that provides that character. By comparing a child’s and adult’s 
version of similar statements, one may get an idea of what some of the tasks 
of socialization are. One can then watch the development of an ‘adult’ by the 
shift in the use of such blank forms as ‘I’m doing X (action) by virtue of Y 
(competent agent).’ One may begin to see over the course of a child’s growth 
that the child will be using himself as a competent agent for an expanding 
class of activities. On the other hand, where someone is a presumptive adult, 
about whom it’s said, “They’re infantile,” we might find that they are 
producing a statement in which X is something that this person could do on 
their own, but the Y proposed is some other agent. 

Or, for example, by examining childrens’ resources we can get a sense of 
what the psychotherapist is talking about when he says that neurotic adults do 
not have a good sense of reality and, again, that they remain children. Let me 
give a conversational excerpt. 

B : How do you make people love you? How do you do it? I wish I knew 

I wish I see people doing it all around me and I try to imitate 
them and I don’t know how you do it. 

A: Uh huh 

B: I don’t know It isn’t because You don’t do it by loving them 

you don’t do it by being thoughtful of them You don’t do it 
by being understanding ... I I really I don’t know how you make 
people love you I just don’t know 

One of the things the woman is doing is asking for instructions: “How do you 
make people love you?” Then she goes through this list: “You don’t do it by 
loving them, you don’t do it by being thoughtful of them, you don’t do it by 
being understanding.” Now where does she come by that list? They’re pretty 
familiar objects; norms. They’re a set of things one is told one ought to do to 
get what she proposes she didn’t get, presumably by using them. 

I want to focus on things like this for awhile; they’re extremely important: 
It’s rather well known that very young children have, from the perspective of 
adults, a rather poor notion of causation. They don’t know how things 
happen to happen. Now, among the ways that adults go about formulating 


Part I 

rules for children, are two which it’s important to distinguish. Call them Class 

1 and Class 2. A prototype of Class 1 is, “Don’t stick your hand on the 
stove.” Prototypic of Class 2 is “Honor thy father and mother’ - and such 
things as “If you want people to love you, you should love them, be 
thoughtful of them, etc.” belong in that class. 

With respect to an adult’s conception of reality we would say that these 
two are different, in that for Class 1 the consequences, whatever they are, 
naturally flow from the act done. If you stick your hand in the fire, you get 
burned. Whereas for Class 2, that’s not so. For a lot of things that you do that 
are said to be wrong or harmful, somebody has to do something to you for 
you to get the negative consequences. You can ‘get away with’ things of the 
Class 2 sort. 

Now it’s supposed, and it seems reasonable, that there’s a stage when 
children don’t know the difference between those two classes. That fact is very 
important for adults because they exploit it heavily. The way they exploit it 
is, they’ll formulate a whole range of what, as adults, we would talk about as 
Class 2-type rules, in terms of Class 1 operations. That is to say, they 
formulate a whole bunch of those rules for which the consequences occur only 
when somebody does something, as though the consequences occur as a 
natural fact of life apart from anybody’s doing anything. They’ll do that even 
though the consequences obviously involve the adult doing something. So 
that parents say to children while giving them a spanking, “I don’t want to 
do this, it just had to be done,” retaining thereby the relevance of Class 1. 

The fact that there are these two separate classes, and that adults do not 
make a large point of discriminating them, but instead assimilate the one to 
the other, formulates a very serious set of problems for children. And that is 
that children are repeatedly faced with the question, “What kind of a rule is 
that, that I’ve been told? Class 1 or Class 2?” There’s no principled way for 
them to find out. They have to proceed case by case. And proceeding case by 
case, they can get into a hell of a lot of trouble. Adults know this and have 
a whole class of proverbs on it, an instance of which is, “If you tell children 
not to stick beans in their ears, they stick beans in their ears.” Children have 
to find out what ‘sticking beans in their ears’ is a case of. Bettelheim, for 
example, reports kids in Chicago who do things like get into a barrel and roll 
down a hill onto a main street where there’s fantastic amounts of traffic, just 
checking out whether it’s so that you can get hurt. And seriously disturbed 
children are those who go about assimilating the whole range of Class 1 
phenomena to Class 2. They then go about checking out the causal properties 
of the world as though they were normative properties in the sense that Class 

2 rules are. 

When children get brought up with a certain amount of assimilation of 
Class 1 and Class 2 rules, of course they cause tremendous trouble for their 
parents and other adults. For one, if the child is about to do things like 
sticking forks in outlets and other such things, then an adult has to be around 
a lot of the time. And if an adult is around a lot of the time, and is saving the 
child from these pains, then we get another sense in which the child is learning 

Lecture 10 


that the rules are indistinguishable. Whereas an adult spanking a child might 
want to retain the relevance of Class 1; here the child is learning that Class 1 
rules are Class 2 rules because it depends on whether an adult is around to 
punish you, that you get hurt or not. 

Now a major device for children, for separating these things out, are things 
like lies and secrets. They begin to discover that there are some things which 
they can violate, that, if the adult doesn’t know, isn’t told, doesn’t find out 
about, nothing happens. And that may operate in alternative ways. One way 
it may operate is that they develop an adequate notion of reality; that is, they 
separate Class 1 and Class 2 rules. An alternative way that it may operate — 
and this is one of the ways that a person who is said to be neurotic can be seen 
to be neurotic - is that one doesn’t see rule violations which have had no 
consequences as involving, therefore, Class 2 rules, but as involving Class 1 
rules with no time-bounds on their operation. And one spends one’s life 
awaiting the natural consequences of those actions which were violative of the 
rules. Persons in such situations live out a great part of their lives under the 
sense of the impending consequences of their violations of Class 2 rules seen 
as Class 1 rules. That’s a considerable part of the ‘neurotic sense of guilt;’ an 
ever-present sense of guilt which consists of their knowledge of the set of Class 
2-type violations that they’ve done, like not loving their parents, etc., for 
which they haven’t - yet - been punished. 

What they tend also to do by not putting a time-bound upon the operation 
of these rules; by not saying if at some time the consequences haven’t occurred 
then they must be Class 2 not Class 1, is to formulate these sets of rules as 
‘prophecies.’ So when their parents had told them “If you don’t change your 
ways no man will ever love you,” then you’ll find forty or fifty year old people 
who eventually can give as an account of the failures of their romances that 
indeed they never changed their ways, and their parents were right. So they 
live out their lives under these rules as prophecies of what it is that will 
happen to them. 

And persons can be found to hold the notion not only that the fact that it 
hasn’t happened yet doesn’t mean that it’s not going to happen, but that that 
only means, for example, that when it does happen it’ll be even more 
ferocious. “Things look great now, but boy, wait till it happens.” And 
positive things are treated, not as contradictions with what your parents told 
you, but, for example, as the rise which will make the fall even more 
dramatic. There are whole sets of paradigms in the history of the culture 
which stand as examples: Sodom, the decline and fall of Rome, etc., where 
the enormous rise can be treated as an example, not of a rise which inevitably 
happened to fall, but the rise that was so great to make the fall worse. There’s 
a thing that the therapists who are lucky enough to have such patients - lucky 
enough because the patients are all rich - call a ‘success syndrome.’ And that 
involves, let’s say, a man who comes from ‘nowhere,’ ‘the wrong side of the 
tracks,’ and at a fairly early age amasses enormous success - in business most 
particularly. And at that point when he’s been elected executive vice-president 
or whatever, he goes into a depression and becomes suicidal. And what it’s 


Part I 

made out to be is the fear that he’s finally arrived at that point where the drop 
will be appropriate to the kinds of sins he’s committed. The sins were such 
that it wouldn’t be sufficient for him to lose some trivial $ 50 -a-week job, but 
now he’s being primed; fattened as it were, for a more appropriate slaughter. 

So that just gives, I think, a sense of what it is that the psychotherapist is 
talking about when he says that neurotic adults do not have a good sense of 
reality. And also when they say that they remain children, and that the 
projected operators in their presumptively adult lives are always parents, you 
can see what it is that the psychotherapist is seeing. 

Lecture 11 

On Exchanging Glances 

An assignment has been given, having to do with observing people exchanging 
glances. The recording starts after whatever description of the assignment has 
been made. 

Q : Can I ask a question about this assignment. 1 

HS: Yes. 

Q : For each person that we notice looking over at somebody are we 

supposed to ask them if they know the person? 

HS: No! Don’t ask! 

Q : But we have to write down the class and everything, how can we 

possibly know if we don’t ask. 

HS: Yeah, class membership doesn’t mean Junior, Senior. 

Q: Oh I realize that. 

HS: I don’t mean social class, either. I mean class in the sense that I’ve 
been talking about class. Any class. Whatever it might be. You 
figure it out. When two persons exchange glances, see if you see 
anything similar between them, and see if you see what it is that 
might be what they’re noticing. If you find that you don’t know at 
all, you can say that. I don’t think you’re that naive. You walk 
through the streets and you’re constantly classifying the persons you 

Q: ((re would it matter if the glancers knew each other; e.g., were sitting 

at the same table and someone walks by and the two at the table 
exchanged the glance.)) 

HS : No, I said I wanted persons who are not interacting with each other. 
Though you can add variations if you please. But the specific 
assignment is for persons who are not otherwise interacting with each 

A combination of three sessions. On tape 6, side 2, prior to the lecture proper, an 
assignment is given and discussed. Then on M4, side 1, part 2 an entire session is 
given to remarks based on the papers handed in by the students. That is continued 
on tape 14, side 1. 

1 This lecture differs from the others in that, among other things, it includes many 
student comments, contributions, and questions. These have been omitted from the 
other lectures as much as possible, materials generated by them having been 
assimilated into the text. 



Part I 

I’ll give you an example of what I mean, just from a lay approach to the 
matter. I was walking down the hall the other day, to give an exam to one 
girl. She was standing, leaning up against the wall. In between us walked 
another girl. She passed this girl first, and then me. And the girl who was 
standing leaning against the wall looked at me and gave a shrug of her 
shoulders with a big smile, which I returned. And I don’t think it was a big 
puzzle over what was going on. The girl who walked by was smoking a pipe. 

Now, the two of us knew what we were noticing. But that can be 
problematic. For example, on the Berkeley campus or in places in Berkeley, 
you often find interracial couples wandering around, one of whom is Negro, 
one of whom is white. And people who look like tourists, visitors to the 
campus, etc. - that is, strangers - will stop to look at these couples, and then 
check out with others around. The question-form might be seen as something 
like “Am I in Rome, or am I here?” That is, Rome, generically. “Do I know 
where I am so that I know that that’s something okay or something odd.” 
And when they do that, people will not infrequently just look back at them 
and give them a negative stare. As if to say, “Who the hell are you.” And 
that’s treated as very disturbing. 

So sometimes when you search out somebody to exchange glances with, it 
may be about something you took to be deviant. And for something deviant 
it may be that you look at anybody. Now, for other noticeables, persons may 
pick out special classes. I want you to try to see what the classes are. I have 
by no means collected these things systematically, and it’s done all the time. 

Q : Something occurs to me. I can usually tell when some sort of 

interaction is occurring, let’s say, and someone is looking at me. But 
if someone is looking at someone else, it’s going to be kind of hard 
to differentiate between people that just happen to be looking at each 
other, as opposed to ones that there’s some reason for an interaction. 

HS: Try it and see. 

Q: and if you make a mistake, you’re liable to come to a lot of false 


HS: Try it and see. If you feel doubtful, put down that you’re doubtful. 
I spend a lot of time watching people watch people and exchange 
glances. And it’s often no big deal to see that people are exchanging 
glances. They turn at each other, across a room or from a distance, 
and exchange big broad smiles sometimes. In some places there’s 
enough turnover so that these exchanges will happen pretty fre- 
quently. And it seems to me that people are noticing all the time. I 
may be all wrong. Maybe you can’t tell when people are looking at 
each other, but you can only tell when people look at you. Maybe 
you couldn’t even tell that. But I would find that odd. 

Q : Do we have to watch others watching each other? or can we watch. 

HS: No, I want you to be watching others watching each other. You can 
add personal remembrances if you want to, or you can begin to 
record the encounters happening to yourself. But as the assignment, 

Lecture 11 


you’re to be watching others looking at each other. And I’m willing 
to make a fair bet that you can guess, seeing somebody get up and 
start to move out of a place, whether they will be somebody that 
others will notice. And you can probably say who will notice them, 
knowing nothing about the persons except what they look like; both 
the person who gets up and the other persons around. Because after 
all, when persons look at somebody passing, they know from having 
scanned the room in the first place, who to turn to to get an exchange 
of glances. 

Q : Is it okay if we write down, if it happened in different places- 

HS: Yes. Put down the place and the time. The more detail the better. 
And if you can type these it would be a help, since I’m perfectly 
willing that anybody in the class takes the collection of them and sees 
what they look like, writes a paper on it, whatever else. We’ll have 
the product of a bunch of persons going out independently. Now 
we’ll see if they look alike. 

I’m going to start to talk about your first assignment. I don’t have all that 
much to say yet. I find them something to think about, and I want to begin 
with something that was suggested by some of the papers. Let me make a 
couple of remarks about the problem of ‘feigning ignorance.’ I found in these 
papers that people will occasionally say things like, “I didn’t really know what 
was going on, but I made the inference that he was looking at her because 
she’s an attractive girl.” So one claims to not really know. And here’s a first 
thought I have. I can fully well understand how you come to say that. It’s part 
of the way in which what’s called your education here gets in the way of your 
doing what you in fact know how to do. And you begin to call things 
‘concepts’ and acts ‘inferences,’ when nothing of the sort is involved. And that 
nothing of the sort is involved, is perfectly clear in that if it were the case that 
you didn’t know what was going on - if you were the usual made up 
observer, the man from Mars - then the question of what you would see 
would be a far more obscure matter than that she was an attractive girl, 
perhaps. How would you go about seeing in the first place that one was 
looking at the other, seeing what they were looking at, and locating those 
features which are perhaps relevant? 

Now the matter is important because there are many occasions on which 
you might want to feign ignorance. I want you to come to see that that has 
to be learned as a special task. Here’s the sort of thing I mean. In a short story 
by Conrad Aiken which I forget the title of, some fellow is engaged to some 
girl who claims that she’s exceedingly innocent. And what bothers him is that 
her body doesn’t seem innocent; that she knows what to do. She knows too 
well how to behave, it seems to him, to have not done elsewhere what she’s 
doing with him. And the task can be quite delicate. For example, sometimes 
women can be in a group of persons one of whom is their husband, and their 


Part I 

husband takes it that they’re not very knowing. And they play that with him, 
but can perform in such a way that others will know that that’s not so. 

Here’s a related sort of issue. I once glanced at a paper concerned with 
teaching actors how to learn to make an error in doing something. For 
example, if your job is to be some professional, and in the play that 
professional makes an error, you have to learn how such a one, who knows 
how to do something, could do that erroneously. And the various set of things 
that you go about doing erroneously may give away the fact that you know 
just what you’re doing. There could be erroneous ways of doing integration 
in calculus, one of which would indicate that you have no idea what those 
symbols are, others of which might indicate that you know how to do 

I think of a classical story, a very fantastic event that happened in the last 
couple of years. There was an automobile accident outside of Moscow. One 
of the victims was the leading theoretical physicist in the Soviet Union, and 
one of the very few leading theoretical physicists in the world. He was killed. 
But they didn’t want him to be killed, and so, through some unbelievable 
exercises they organized a staff from around the whole world to bring him 
back. And they reassembled him, essentially. He was absolutely destroyed; 
clinically died five times in the first ten days, and they had a fantastic team of 
persons coming in to put him together again. Now the question was, was he 
going to be the same guy? So as he began to recover - and that meant very 
dramatic kinds of developments from essentially infantilism - his colleagues 
would go in and talk to him. 

One crucial day, one of the colleagues comes out very depressed and says, 
“He’s not himself.” And they said “Why?” He says “Well, we were having 
this discussion about some problems and I asked him a question, and he came 
back with this answer, and that’s just not so, and anybody knows it’s not so.” 
And they all sat there bemoaning the fact that it looks like he’s finished. Then 
it suddenly occurred to somebody, “My God, look at that answer!” It was a 
far more elegant solution to the problem than anybody had thought of. And 
it was just not available to any of them, but they knew that that’s the way he 
would work. So that situation was transformed very dramatically. And in that 
regard, let me just note that, for example, things that children say, which they 
have no right to say, can be treated as ‘errors.’ That is, errors are socially 
stratified in many ways. 

Let me go at this business of 'inferences’ in another way. The problem that 
I am stuck on at this point, and I don’t have anything like a solution to it, is 
how does a glance become an action? What kind of a world do you have to 
build to make a glance an action? Let me start off by reading you a quote from 
an extremely important book, the title of which I also forget. I think it’s called 
Sight and Sense , but you can easily find it given that the author’s name, Von 
Senden, is not that common, and he only has one book translated into 
English. It was written in, I guess, nineteen twenty something. It by and large 
reports on others’ investigations which had never been brought together. The 
book is exceedingly important in a lot of ways. The very eminent psycholo- 

Lecture 11 


gist, Head, made a great deal of his reputation on his analysis of the materials 
Von Senden presents, which Head was able to do because he knew German 
and the book hadn’t yet been translated into English. And the theory that he 
developed is built largely on Von Senden’s work. So far as I know, sociologists 
have not used the book; the place where you would expect to find it among 
the classical sociologists is Sandeke, but I don’t see that he even had any 
reference to that kind of material. 

The book is about persons who, typically, were blind from birth and who 
got sight, and what kinds of things they could learn to observe afterwards, 
and what kinds of things they never learned to observe if they got sight too 
late. On pages 61-62 we get a report by Dufau of “a girl who only 
discovered at the age of 12 that she differs from other people in lacking a 
sense, and who now seeks to discover the nature of this unknown sense.’’ And 
she says: 

I posed myself a host of questions about this new and unknown state 
which had been described to me, and did my best to come to terms with 
them. In order to satisfy my doubts I had the idea of trying a strange 

One morning I again put on the dress which I had not worn for 
some time because I had been growing so rapidly then from month to 
month, and thus attired I suddenly showed myself at the door of the 
entry room in which my governess was already working at the window. 

I stood listening. “Good heavens Lucy,” she said, “why have you put 
on that old dress that only reaches to your knees?” I merely uttered a 
few idle words and withdrew. This was enough to convince me that, 
without laying a hand on me, Martha had immediately been able to 
recognize that I had again put on the dress that was too short. So this 
was seeing. 

I gradually recounted in my memory a multitude of things which 
must have been daily seen in the same fashion by the people about me, 
and which could not have been known to them in any other way. I do 
not in the least understand how this happened, but I was at last 
persuaded. And this led gradually to a complete transformation of my 
ideas. I admitted to myself that there was in fact a highly important 
difference of organization between myself and other people. 

Whereas I could make contact with them by touching and hearing, 
they were bound to me through an unknown sense, which entirely 
surrounded me, even from a distance, followed me about, penetrated 
through me, somehow had me in its power from morning to night. 
What a strange power this was, to which I was subjected against my 
will, without for my part being able to exercise it over anyone at all. It 
made me shy and uneasy, to begin with. I felt envious about it. It 
seemed to raise an impenetrable screen between society and myself. I felt 
unwillingly compelled to regard myself as an exceptional being that 
had, as it were, to hide itself in order to live. 


Part I 

Most of Von Senden’s work concerns the perception of patterns. If persons 
are blind from birth and they get their sight as adults, it turns out that certain 
patterns just can’t ever be learned. They may not ever learn how to see a 
rectangle. They know what a rectangle is, and they can find it by counting the 
points. But where one sees, quote, at a glance, a rectangle, a triangle, and 
other such things, they never learned how to do it. Now there are lots of other 
things that people see, quote, at a glance - and perhaps that blind persons 
know the categories of. And how it is that persons come to see these things 
that they do see in this fashion we don’t know very much on. But what’s 
important in the first place is to try to determine, in part, what it is that is seen 
in this fashion. 

And that’s why it’s sort of a pain to intellectualize this stuff such that you 
already talk about it as though, “I see a blob, and then I infer that it’s my 
mother because she’s a blob like that,” when what you see is not that. And 
it’s an extraordinary experience when it turns out that you do see somebody 
in something like that fashion. I just recalled, in Fitzgerald’s novel The hast 
Tycoon, he reports a scene where this girl remarks that walking in New York, 
seeing a man approaching, finding a whole series of properties of him which 
she doesn’t especially like, it turned out it’s her father. That is a special sort 
of experience, but I can understand why you talk that way. You may find that 
some philosopher has convinced you of this. But ‘inference’ has another kind 
of use. 

So where are you going to start to try to build a way of dealing with this 
stuff? We start out with the fact that glances are actions. That’s the first fact. 
There’s a beautiful report to this effect. A guy is looking at a girl, looks 
around to find somebody to exchange glances with, catches the eyes of another 
girl who looks like this one in some way, and turns away quickly. That is, he 
sees her sanctioning his looking. And how do we start to provide for glances 
as actions? I take it we have to start building classes. Earlier I was talking 
about persons being representatives of classes of which they were members. 
And the sorts of things you’ve reported suggest that what persons see is a class 
member, for whole collections of classes. You see ‘a girl,’ ‘a Negro,’ ‘a 
such-and-such.’ That is, the class permits you to see what it is that’s there. It 
permits you to see. 

To liven this matter up, I’ll read you something from The City of Plains 
by Proust. And Proust is an incredible sociologist, as you may know if 
you’ve read it - and if not you certainly ought to, even if you’re not 
interested in literature. There is a scene where Proust is watching events 
take place in a courtyard below. He sees a whole sexual confrontation 
between two guys, which he describes absolutely fabulously. Then he 

From the beginning of this scene an evolution in my unsealed eyes had 
occurred in M. Charleaux. As complete, as immediate as if he had been 
touched by a magician’s wand. Until then, because I had not under- 
stood, I had not seen. 

Lecture 1 1 


The vice - we use the word for convenience only — the vice of each of 
us accompanies him through life after the manner of the familiar genius 
who was invisible to men as long as they were unaware of its presence. 
Our goodness, our meanness, our name, our social relations do not 
disclose themselves to the eye. We carry them hidden within us. Even 
Ulysses did not at once recognize Athena. But the Gods are immediately 
perceptible to one another, as quickly as like to like. And so, too, had 
M. de Charlieux been to Chiupien. 

Until that moment I had been, in the presence of M. de Charlieux, 
in the position of an absent-minded man who, standing before a 
pregnant woman whose distended outline he has failed to remark, 
persists, while she smilingly reiterates “Yes, I’m just a little tired now,” 
in asking her indiscreetly, “Why? What is the matter with you?” But let 
someone say to him “She is expecting a child,” suddenly he catches 
sight of her abdomen and ceases to see anything else. It is the 
explanation that opens our eyes. 

So, the classes and their categories permit you to see. That’s a start. It’s not 
enough to make a glance an action. As some of you noticed and reported, it’s 
not merely that some observer is seeing by reference to some category, but that 
the one being observed sees what the observer is, and is seeing. And we get 
into that whole jumble there: A, seeing that B is looking at A, sees what B is, 
and what B sees, etc., etc. 

That A can see what B is, and what B is seeing, may seem the most trivially 
obvious fact. But it was the economist Bagnanin’s seeing that fact that 
provided for the modern revolution in economics. Until the book A Theory of 
Games in Economic Behavior, the major theory of economics used the Robinson 
Crusoe model: A man alone in an environment, and now he has to go about 
deciding what to do, what things will work, etc. And any other person is to 
be conceived as a part of the environment. That meant that you could give 
statistical treatment to the various parts of the environment. Now Bagnanin 
was, among other things, a great poker player. And he saw that economics 
could not be constructed along the Robinson Crusoe model, and took poker 
as a model. And in poker you can’t treat the other person as a statistical object, 
but as somebody who, whatever strategies you might employ to deal with 
‘that piece of the environment,’ does the same about you. And then, 
furthermore, knows you do use strategy, etc., etc. If you just read the first 
chapter of the book, you see that laid out. As I say, it provides for a complete 
reconstruction of the way of doing economic theorizing. It’s a very curious bit 
of history that Parsons, in his Structure of Social Action, posed essentially the 
same facts - and he left economics because of a similar complaint against 
economics. But the sociologists apparently didn’t appreciate what Parsons was 
posing, and have come to this kind of position through Bagnanin and 
Morgenstem, and not through sociology itself. 

So, A can see what B is, and what B is seeing. How is it that, given there’s 
a whole bunch of classes available, A can see that? I have to introduce a notion 


Part / 

that will help this problem along. Let’s talk about there being, for some 
collection of classes, an ‘order of relevancy’ with respect to categories. It seems 
that a set of circumstances can provide that order of relevancy for some 
membership class. If the circumstance is that A is being looked at by B, that 
in itself might inform the consideration of what order of relevancy is operating 
here. And if there is an order of relevancies, we can begin to locate certain facts 
that obviously do occur. For example, ‘failures’ can be located that way. And 
‘absences.’ People talk about “missing something,’’ “not seeing something.” 

Now a third thing may be very extendable, but for now I’ll only deal 
with classes that have two subclasses. For some classes. A, being observed, 
sees that B, doing the observation, sees that A is a member of some contrast 
class such that A is one part of a paired class and B is the other. Where for 
each, there is an order of relevance which provides for the other as being an 
observable, and - and this is absolutely important and the thing would not 
work unless it was present - the order of relevance of each was available to 
the other. 

With that, we can begin to handle some rather nice kinds of events; you 
can begin to see what the following kind of trouble is. A girl looks at a guy. 
And what he does is, he takes her glance and then looks over his shoulder. 
And what’s going on is that he knows enough to know the sort of thing she’s 
looking at, by reference to the class - she’s looking at ‘an attractive 
male’ - and he has that sort of insecurity which provides that he’s not eligible, 
so it must be somebody else. This complementarity of orders of relevance 
permits us to see how persons are operating when they talk of themselves as 
‘worthless’ or as ‘nobody.’ And it’s this complimentarity - that one knows 
what the other classes do - that begins to tell us what may be going on, in 
part, where some guy is looking at a girl, now looks around and catches the 
eye of another girl who is somewhat similar to the first, and turns away. That 
is, he does not seem to just continue looking around as though, “Well, she’s 
not a male so how would she know what I was doing?” Or, for example, some 
of you report a person will be walking, sitting studying, just doing nothing, 
and then see somebody looking somewhere, and, quote, follow the glance, 
knowing that there would have to be something at the end of the glance 
which is worth looking at. And if you can see what it is that is doing that 
looking, you could have a pretty good idea of what it is that would be at the 
end of it. So this complementarity is equally as crucial as the fact that one is 
able to see what somebody with whom you are a member of a class in 
common is seeing when they look at you, or another. The sense of there being 
‘a society’ is that there are many whomsoevers, who are not members of this 
or that class, who are able to see what it is that one is looking at. 

Now there are other things that you talked about, which are worth 
thinking about. For example, people talked about the fact that persons didn’t 
exchange glances over an average person. Now what you have to attend is the 
notion of ‘average’ that you employ. Because I don’t think it’s the case that 
you use it in the following way: You’re sitting someplace, a set of persons pass 
you by, and you construct a distribution over that set of persons and provide 

Lecture 1 1 


that those that stand out from that set, after let’s say, watching awhile, are the 
unaverage ones. Rather, somebody’s being outstanding may be quite irrele- 
vant to the collection of persons surrounding them in the scene; that is to say, 
those who have passed before or who come after. 

A while ago I gave some examples of odd suicides, and I said then that 
given the procedure that’s used - that is, case by case - what you have is that 
an odd suicide gets separated off and is not used to consider the, quote, 
normal ones. 2 So a stream of odd ones is only something to be remarked 
upon, like “Isn’t it amazing today, there’s such a stream of absolutely 
beautiful women who walked through such-and-such a place.” It isn’t that 
one then finds that one is going to modify the notion of ‘average’ because of 
that. And that then means that those categories have to be given special 

In many of your reports it’s proposed that what makes something 
noticeable is that it’s an ‘incongruity.’ I’ve given a glancing attention to the 
phenomenon of incongruity for a long time, but it’s only now that I feel that 
I’m beginning to get a handle on what I might be able to say about it. 

A first thing we need is a notion that what Members see is decomposable 
by them. That Members can decompose some event, situation, complex, 
whatever you want to call it, is no surprise at all, given the sorts of things 
we’ve been considering. That is to say, we’ve been talking about activities as 
being ‘assembled.’ And if that’s so, and if Members can see that that’s so, 
then, that they can take them apart would not be especially surprizing. The 
question then is, having taken something apart, how do they put it back 
together again so as to find what it is that’s strange. 

The way they seem to do this involves treating something that they see as 
a combination of parts, some of which have names. And to those nameable 
parts are affiliated standardized procedures for producing those objects in 
some combination. I’ll give you an example. You see a man in a car. Now a 
car is a ‘possession,’ and there are ways that one properly, normally, 
legitimately, expectably comes to have it. And apparently what goes on is that 
when one sees ‘an incongruity,’ one of the things one sees is that it does not 
seem that the proper procedure for this possession having come to be 
possessed, has produced the combination of this person and that car. 

A guy I know told me the following story. He looks like a bum, is usually 
unshaven, wears very tattered clothes, and has a big flashy car. He’s driving 
down the Massachusetts Turnpike and there’s a police car sitting in the grass 
that divides the highway. He passes the police car, drives along below the 
speed limit, the police car starts to follow him, follows him along for a while, 
finally pulls him over. The policeman asks him for his license, the guy shows 
him his license. And now we get the following interchange. “What do you 

2 See lecture 8, p. 62. 


Part I 

do?” “I go to school.” “Where do you go to school?” “Harvard.” “That’s 
very nice. Where are you going?” “I’m going home.” “Whose car is it?” “My 
mother’s.” Then the policeman says, “Look, would you do us a favor?” 
“Sure. What do you want.” “Would you get a shave? if you don’t, every 
policeman on the Turnpike is going to have to stop you.” That is to say, for 
such a person, how it is that they come to have gotten that car is a problem. 
And the incongruity is seen in that way; by virtue of the car being a 
possession, how would it come to have been possessed by such a one? 

Or, for example, I talked about a woman who proposed that she was 
‘nothing,’ and said “What man would want me?,” given her set of features. 
And in analyzing that, I said that what she was proposing was that given her 
features, the correct procedures for somebody selecting a woman would not 
arrive at her being in some pair. 3 And we can see a sense in which, whether 
in any given circumstance the person who says that is correct or not, they are 
in fact employing a procedure which might be employed if they happened to 
be together with someone; that is, people would say “How did she come to 
get him?” That is to say, one thing which seems to be observable as an 
‘incongruity’ involves two persons being present, a man and a woman, where 
there’s some special difference between them: A worn old man and a very 
young pretty girl, or the reverse. And now they get noticed, an exchange of 
glances takes place, and the question that seems to be asked is “How did he 
get her?” or the reverse. What one is doing is employing the procedure by 
which persons properly come together and finding that that does not produce 
these two persons as a pair. And one can then produce an explanation; for 
example, “He bought it” or “she bought it.” 

So that the procedures whereby persons come to be in some combination, 
or come to have some object, seem to describe in part what it is that you’ve 
observed, and what I had observed earlier but had never really been able to 
figure out - how it is one sees an incongruity, and also sees the possible 
illegitimacy of some combination. 

Now, in some of the reports what we seem further to find is that where 
persons are concerned, units of larger-than-a-member are observable. People 
can see ‘a family,’ for example. That may be trivially obvious, but it’s very 
important, and it has to be achieved in some way. One of the reports has a 
powerful instance of the relevance of persons seeing such a thing as a ‘family.’ 
One of you reports that you were driving along and a car pulls up and stops. 
It doesn’t seem to stop anywhere special, just pulls up and stops on a street. 
The door opens and a girl of about 18 charges out, runs across a lawn and 
stops, and starts shrieking. In the front seat are an older man and woman. The 
guy jumps out of the car, charges across the lawn, comes up to the girl and 
gives her a smack right in the face. At which point some of the passing cars 
slam on their brakes, and some people start getting out of their cars. 

The man and the girl stand there, face to face, screaming at each other, and 
then he just grabs her and drags her back to the car. And people look at each 

3 See lecture 9, pp. 68-9. 

Lecture 11 


other, shrug, and say “Well,” get back in their cars and go on their way, 
taking it that it’s not after all a kidnapping scene or an attempted murder, but 
it’s ‘a man and his daughter’ and he’s punishing her for something she did, 
or something to that effect. No attempt at intervention is made. Whereas I 
take it that were the unit ‘family’ not available to them, and relevant features 
of it not observable in the scene, then it could be expected that people would 
intervene. But given what they saw, intervention becomes ‘meddling.’ And 
one wonders, again in part, how much of the failure to give help in dramatic 
scenes is a product of some order of relationships where the combination of 
persons and what they’re doing is something seeable as ‘their business.’ 

Ql: I read in the paper where a girl was in a car and she had parked a 

couple of blocks from her home. And another car pulled up and the 
guy driving asked her for information, “How do you get to the 
freeway?” She started to tell him, still being seated in the car. The 
guy said “I can’t see you, come around and show me on the map I 
have with me.” And she immediately got frightened, rolled up the 
windows and stayed in there. And he said “That won’t help you,” 
and he tried to get in. And she started honking her horn, hoping that 
someone would come out, hearing the horn. And so what he did was 
honk his born back. And therefore it sounded more like a game, and 
no one came to her aid. 

Q2 : The was another situation that was in the paper last week where a 
woman was sitting in the car waiting for her daughter and her date 
to come out from a show. And this man got into the car and started 
beating her up. And she yelled for help, and there was a couple 
watching, and another man. And no one would help. And after- 
wards, when the guy left, she said “Why didn’t you help me?” And 
the woman in the couple said “Well we thought you were married 
and it wasn’t any of our business.” 

Q3 : But then on the other hand there was a story in the paper about this 

young motorcyclist who stopped on the freeway, who looked down 
below, evidently at another freeway portion where a car had stalled. 
And he leaned over the railing to ask if this man needed any help, 
and the answer was a shot from a gun, and he got killed. So these 
are the kinds of things that may be part of people’s fear. 

Q4 : Which all points to a kind of a pattern of three things: A normal way 

of going about things, a subversive way of going about things with 
a normal appearance, and an uncertainty feature operating with 
regard to all events, since it’s known that persons can be acting in a 
subversive fashion and appear to be acting normally, at the same 
time it’s not known whether in fact they happen to be operating in 
a subversive fashion. And as a matter of fact it would be more of a 
problem to account for why that uncertainty factor is not operating, 
than to account for why it is operating when it is. 

HS : Well, that may be. The problem about what you say is, in part, that 


Part I 

in this society there is a considerable shift of the responsibility for 
locating subversives, from whomsoever to a special group, the 
police. And persons otherwise feel that they’re entitled to suppose 
that people are what they look like. And if they’re not, then that’s 
not their problem very much. It’s fairly special that people use a 
supposition that somebody is not what they seem to be. 

Q4 : It’s Goffman’s argument, the notion that persons feel that if some- 
body is acting in a particular fashion which they may doubt as being 
an example of such-and-such, then they’re entitled to inspect him for 
other features that they assign to a such-and-such. I think that’s much 
more common than the other. Because they’re engaged in the prob- 
lem of attempting to make out persons and what they’re like, and if 
they now do something that you didn’t figure on, it’s obvious that 
they did something. I mean they did it. So it’s obvious that 

HS : Yeah but the fact that they did it is not sufficient. The first thing that 
we find is that events are decomposable. And the question is, “Could 
somebody have done that?” and it could be decided that they 
couldn’t have done it properly, and now we get all these incongruity 
observations. Now the fact is, Goffman talks about incongruity but 
he does not tell us what incongruity is. That’s what I think I’m 
beginning to see here in this stuff. How it is that one sees it. He has 
not analyzed how it is that you do ‘an incongruity,’ what makes it 
an incongruity. And l think I have the beginning of how it is that 
you do it. 

Now such issues as its import for integration in the social structure, which 
is what you’re talking about when you raise ‘uncertainty,’ is another matter 
altogether. Not to say that it’s not worth looking at. And the question of 
uncertainty is in part handleable in the observations you’ve made in these 
reports; that is, that people check out the things they’ve noticed with 
somebody else. What we have to consider is, what is it that they get out of 
checking out their noticings with somebody else? Some of you talk about 
‘reinforcement;’ one reinforces their determination that someone was doing 
something wrong, or that a combination was wrong, by making a check-out 
with somebody else. And thereby the norms are reinforced. Which is a nice, 
Durkheimian kind of argument, and it may be true. But there’s something 
that has to be seen about that claim, and that is the mechanism of the 
procedure in the first place, which is what I’m interested in. 

In the first place, if one were engaging in some device for getting 
reinforcement, it might be supposed that one would want to know the status 
of the person one uses to get reinforcement from. That is, if you have a 
question, “Is it the case that this thing I’ve seen is a violation of the norm?” 
you might write a letter to Dear Abbey, or ask a priest, or ask your parents, 
or ask somebody who has special rights. In any case, it might be presumed 
that you’d be concerned to check with somebody who you knew to have some 
information about it. But in your reports, apparently what we find is that it’s 

Lecture 11 


pretty much an ‘anybody’ who can be turned to for that check-out; ‘anybody’ 
as long as they’re a member of an appropriate class, so let’s say, a woman 
turns to a woman, or a man turns to a man. You don’t ask for their credentials 
in the first place, and if they return your glance, give you the smile back that 
you give to them, or the disapproving glance, that seems to be okay. And 
that’s a fantastic kind of simplicity. 

And lurthermore, it might be supposed if one was getting a reinforcement, 
that if the person did not return the same glance, give the same thing back to 
you, then you would say “Maybe I’m wrong,” figure that you’re having a 
wrong response. But no, that doesn’t seem to be the case. Given what you’ve 
reported in your papers, what seems to be the case are several things. First, if 
they don’t return the same glance, a smile for a smile, etc., you look for 
somebody else, whereas if they do return the glance, you feel satisfied and look 
no further. And second, if they don’t return the glance, you may make an 
assessment about them: “Who do they think they are?” or “What’s up with 

So the norms don’t seem to be in doubt; that is, the norm that provided 
for the incongruity that provided for the noticing is not something that seems 
to be held in doubt. And the notion that it’s ‘reinforcement’ that you get is 
somewhat obscure. Though the question certainly has to be asked, what is it 
that people are doing when they exchange these glances? We’ve already 
observed that they get exchanged with whomsoever; there’s no special 
credentials involved, except class membership. Another, equally extraordinary 
part of this - and why, again, ‘reinforcement’ is a tremendous gloss or 
oversimplification - is that one takes it that they know what it is that you saw, 
that they saw the same thing, and they know what you’re smiling about, and 
they make the same assessment. Their smiles tell you that. Now, that that 
gets done would seem to involve a fantastic kind of social integration. And 
it can give us a way to talk about ‘alienation.’ Because one of the senses of 
alienation would presumably be that you would feel tremendous doubts 
about doing that. And one would have tremendous doubts about what 
those smiles and disapproving glances are about; what they’re looking at 
you for. Schizophrenics are always reporting that people are looking into 
them, talking about them, know what they’re thinking. And you can see 
that they haven’t constructed a machine which produces events going on in 
the world, wholely out of their heads. For that thing which they propose is 
going on by reference to them, while it may not be going on by reference to 
them, is indeed going on, enormously, and how it’s going on is an absolute 

The sense of alienation with respect to a schizophrenic would be, in part, 
that they can’t tell when it’s being used with respect to them. Or, that they 
think it works better than it works. Because it’s perfectly routinely the case 
that persons know what others are thinking - without knowing those 
others - in just the sense that ‘thinking’ is used as a non-technical term. 
People seem to know what others are thinking without having any idea who 
they are, apart from their class membership. Now schizophrenics claim that 


Part I 

others, who don’t know them, know what they’re thinking. Routinely. Well, 
there’s a sense in which others do know what others are thinking, routinely. 
Again, then, a schizophrenic may be wrong as to when it occurred, but the 
notion they have is obviously correct; is obviously so. 

Q: It seems like that was the basic assumption of the assignment; that 

we could notice them doing it. 

HS: Right. It was the basic assumption, which could have been wrong. 
But of course it’s not like I came into this world the day before I gave 
that assignment. I know that people can do this, I’ve watched it 
many times, and I take it that you’ve seen it also. However, people 
who get this far in their education are very prone to intellectualize 
the whole operation and not see it. And so you don’t see that people 
know what other people are thinking; you figure it’s a philosophical 
impossibility, or you’d have to go through four years of analysis, etc. 
It was an assumption. And it could have been the case that 
everybody came back and said ‘‘No, I never saw that happen.” And 
that’s possible. It might be something that’s dying out. A thing that 
our forefathers had. Like God. 

Lecture 12 

Sequencing: Utterances, Jokes, 
and Questions 

For the linguists, almost exclusively the largest unit of investigation, the 
largest unit they seek to describe, is a sentence. So grammar is directed to 
providing rules for generating sentences, and every time you have a different 
sentence the grammar is to be reapplied. If we want to study natural activities 
in their natural sequences, we have to deal with, for example, the obvious fact 
that a sentence is not necessarily a ‘complete utterance.’ Thus, linguistics is not 
sufficient, at least so far as it’s by and large done. There is one major exception 
and it’s extremely close to what I’m trying to do. That is Fries’ book The 
Structure of English (1952). 

We want to construct some unit which will permit us to study actual 
activities. Can we construct ‘the conversation’ as such a unit? Can we in the 
first place make of it ‘a unit’ - a natural unit and an analytic unit at the same 
time? The question then becomes, what do we need, to do that? 

First we need some rules of sequencing, and then some objects that will be 
handled by the rules of sequencing. Now, if we restrict our attention at the 
beginning to two-party conversations, then we can get something extremely 
simple - though not trivial, I assure you. And that is, that for two-party 
conversations the basic sequencing format is A-B Reduplicated. It’s not trivial 
in that with three-party conversations it’s not the case that the sequencing 
rules are A-B-C Reduplicated. There’s something else; what it is, I don’t 
know. So: A-B Reduplicated. One party talks, then the other party talks, then 
the first party talks again, etc. I use the term ‘two-party’ so as to provide for 
the fact that this does not necessarily mean two persons. The ‘two-party’ 
conversation may be a basic format such that conversations having more than 
two persons present can take a two-party form. That would involve persons 
dividing themselves up into teams of a sort, and alternating according to team 
membership, where, then, one team talks - a whole series of persons might 
talk for that team - then the other team, etc. 

Restricting our proper considerations to two-party conversation with the 

The nine Winter 1965 lectures (all of them owed to ‘transcriber unknown’) pretty 
much recapitulate the considerations of ‘Fall 1964,’ sometimes in a more developed, 
formal way. Most of them have been incorporated into those earlier lectures: lecture 
(1) has been absorbed into lecture 1; lectures (4) and (5) into lecture 6, lecture (6) 
into lecture 9, much of lecture (7) into lecture 3, leaving lectures (2), (3), (8), and 
(9). lectures (2) and (3) comprise this lecture 12, (8) and (9) comprise lecture 13. 



Part I 

sequencing rule A-B Reduplicated, what we have to come up with as a first 
object is something we can call an ‘ adequate complete utterance.' And that will 
be something that a person can say, which, upon its completion, provides for 
the relevance of the sequencing rules. That is, on its completion, the other 
talks, properly. A sentence may be complete, and one could tell that it’s 
complete, but that wouldn’t tell you that the person is finished speaking for 
now. But if they use an 'adequate complete utterance’, then, by virtue of the 
fact that that unit is complete, the sequencing rules are relevant. Again, a 
sentence is in general not sufficient, though some sentences may be as much 
as a person is going to say. Nor, for example, is it generically the case that a 
‘question’ is sufficient, though questions may comprise, and frequently do 
comprise, complete utterances. That is to say, it is not enough to propose that 
an item was ‘a question,’ to know that upon its completion the other was to 
have talked, or did talk. 

We want to see if we can get something that stands as an ‘adequate complete 
utterance,’ such that upon the use of one or several of these, we have minimally 
constituted something that will be, recognizably, ‘a conversation.’ This is not to 
say that conversations are only built up out of adequate complete utterances, 
because persons can have ways of detecting that something is or is not com- 
plete, apart from the fact that the object is standardized in such a way. Though 
insofar as they’re using other things than adequate complete utterances to 
make up their talk, they have a special task of detecting that the other is or is 
not finished, and perhaps what they ought to do now. 

Another way that we could tell that ‘a conversation’ has taken place is if 
there were some invariable part. We might then go about identifying the fact 
that a conversation occurred by reference to the fact that the invariable part 
occurred. Only, as far as I can tell, there aren’t any. However, there is 
something pretty close to that. Take something like “Hello,” “Hello.” Now, 
a ‘greeting’ is an ‘adequate complete utterance.’ It’s standardized as such. 
When you hear ‘a greeting,’ then you can take it that when it’s complete, it’s 
your chance to talk, if you’re the one that’s been greeted. There are several 
things we want to notice about greetings, apart from the fact that they are 
adequate complete utterances. Greetings are paired. And by that I mean 
simply that if A picks a member of one of those things, then a proper move 
for B when he has an opportunity to speak - right after it - is to pick a 
member also; the same, or another. So one party’s use of a greeting provides 
for that minimal exchange, “Hello,” “Hello.” 

Now, it’s the case that if A-B Reduplicated is the format of conversations, 
then there is no specific length that a conversation takes, to be ‘a conversation.’ 
And there may be no generic way built into the rules of conversational 
sequencing, that a conversation comes to a close. So, for example, there can 
be enormous variance between two conversations as to how much was said. 
That is, you don’t have a situation where some certain amount of talk is 
required before the conversation can, or ought to, close. Or, for example, 
there can be an enormous variance as to how much one person has said, as 
compared to the other. It’s not a situation where persons have to monitor how 

Lecture 12 97 

much they’ve talked as compared to how much the other has talked, to find 
that the conversation can, or ought to, close. 

Thus - if two things were so which are not so - we could say that we have 
a ‘minimal conversation,’ “Hello,” “Hello.” And we could say that if at least 
that took place, then a conversation occurred. And we could describe how that 
could take place, given this A-B Reduplicated format, given an ‘adequate 
complete utterance,’ given the ‘paired’ characteristic, plus a few minor things 
which I’ll point to later on. But the two things are not so. One is, it’s not 
invariably the case that things we would say are ‘conversations’ contain 
greetings. The second is, it’s not invariably the case that ‘greeting items,’ such 
as “Hello,” occur as ‘greetings.’ 

Now those facts lead us to require the following: We need to distinguish 
between a ‘greeting item’ and a ‘greeting place.’ Where, then, something is 
a ‘greeting’ only if it’s a ‘greeting item’ occurring in a ‘greeting place.’ If a 
greeting item occurs elsewhere it’s not a greeting, and if some other item 
occurs in a greeting place it’s not a greeting - though some items that are close 
to greetings might take on the character of a greeting by occurring in a 
greeting place. We need, then, to be able to say that there’s a ‘greeting place,’ 
and that any ‘conversation’ has it. And I take it we can say that there is a 
greeting place in any conversation, by virtue of the following kind of 

First of all, it does seem that there is no rule of exclusion for greetings. 
People can know each other 35 years, talk to each other every day, and 
nonetheless greet each other when they begin a conversation. But take a whole 
range of other items, for example ‘introductions’ (telling someone your name, 
etc.). About introductions it can be said that there are rules for their historical 
use. At some point in the history of persons’ conversations, introductions are 
no longer relevant. And if they’re not relevant, then, when they don’t occur, 
one can’t say they’re not there because there’s no reason to suppose that they 
would be there. Notice that what we re trying to do is find some way of 
saying, non-trivially, that something is ‘absent.’ If there were something that 
was invariably present we would have no trouble. We could say if that thing 
happens, then ‘conversation’ occurs. But we need to be able to say that we 
have a conversation if that thing is present, or if it’s absent. And to say that 
something is ‘absent’ is a much harder task. The way we can go about doing 
that is to find that it’s always relevant. If it’s relevant, then if it’s not present, 
we can say non-trivially that it’s not there. And greetings have that sort of 
relevance, in that there is no rule of exclusion for them. So we can say that 
greetings are relevant for any conversation. 

Secondly, we can distinguish between the greeting item in the greeting 
place, and the greeting item elsewhere. That is, somebody can say “Hello” in 
the middle of a telephone conversation, where what they’re doing is not 
‘greeting,’ but checking out whether the other person is still on the phone, and 
a variety of other things like that. Thus, the greeting item, to be ‘a greeting,’ 
has to occur somewhere in particular, and we can say, then, that there’s ‘a 
place.’ And in that greetings are relevant for any conversation, that place in 


Part 1 

which they are recognizably not something else, but ‘greetings,’ is present for 
any conversation, whether there is a greeting item in it or not. 

It’s not, then, that we just need “Hello,” “Hello” or members of that class 
to have taken place, to have ‘a conversation’ and to warrant our being able to 
say that there is a natural analytic unit, ‘conversation.’ But if we can say about 
some piece of talk - either a greeting item or a greeting substitute such as 
“How are you?” - that it occurred in ‘the greeting place,’ and that piece of 
talk, whatever it was, provided for the relevance of the sequencing rules, then 
we could say that we have ‘a conversation.’ And then we would be able to 
warrant, at least in part, the fact that there is a unit, ‘conversation,’ which is 
natural and analytic, and is generically usable. Of course the description of 
cases of it may be far more complicated than this, but it has to be noted in the 
first place that such a warrant has never been made, and that such a unit has 
not been established. And it’s for that reason, at least, that it’s of interest. 

Now, it does seem to be the case that “Hello,” “Hello” is a ‘minimal 
conversation.’ Persons take it that it’s a minimal conversation. It’s not 
sub-minimal; you don’t need more to have had ‘a conversation.’ And you do 
need that, or substitutes for it. Children learn this at a rather young age, and 
you find them producing perfectly recognizable ways of indicating that they 
have engaged in the beginning of what may be only a minimal conversation, 
and have not been properly treated. That is to say, one of the ways that one 
shows that one has done something which is an adequate complete utterance 
- that is, which is appropriate for the use of the sequencing rules - is to repeat 
it. So I have these reports where a child says “Hi,” there’s no answer, and the 
child says again, “Hi!” And then there’s a “Hi” in return, and the child will 
take that as having been sufficient, and go about his business - which he 
doesn’t do when he says “Hi” and there’s no return. 

That use of repetition as a way of indicating in the first place that an item 
was adequate for whatever it is that’s supposed to come next, is obviously the 
simplest way of doing that task. In the child’s use of it, however, we get 
something that’s worthy of some brief mention. And that is the way that 
adults come to see that the child knows something of language. The way 
adults know that the child is now ‘speaking’ and not babbling, involves the 
fact that the minimally recognizable units of infant speech - and this is 
essentially cross-culturally valid - are combinations of ‘p’ or ‘t’ or ‘d’ followed 
by a vowel like ‘a’. And those combinations seem by and large to be used 
without respect to what the language is; that is, without respect to however 
the adult language may be constructed. And the way, apparently, that one 
tells that the child is now speaking is by virtue of the fact that it doesn’t 
simply produce a series of syllables, but it repeats a syllable. In this culture, 
then, prototypically the first word that a child speaks is the word for ‘father,’ 
“da da .” 1 The interest of this phenomenon lies in the fact that if you get that 

1 Sacks cites the paper in which “this is all discussed,” but the title wasn’t caught 
by the transcriber. One possibility is Ruth Hirsh Weir, Language in the Crib (The 
Hague: Mouton, 1962). 

Lecture 12 99 

kind of stability, across fantastically different languages, then the social 
sciences and biological sciences come to some close relationship. 

So, in the first place, this duplication business is a non-trivial fact, and it’s 
pretty much as simple a way as you can have of indicating a range of things 

- in this case, that something had been done, and was adequate for the 
relevance of the sequencing rules. And you can notice the way that parents 
point out to children their violations on this matter. Suppose somebody comes 
to the house and says “Hi” to the child and the child doesn’t respond. One 
thing the parents will say is, “Didn’t you hear them say ‘hi’ to you?” Where 
they take it that they don’t have to restate the sequencing rules, but simply 
point out that the rules have been adequately invoked. 

1 take it we can say, then, that the unit ‘conversation’ is warranted by the 
fact that we have at least a minimal thing that’s recognizable as ‘a 
conversation.’ For it, the sequencing rules are relevant. We can talk about 
places in it, or a place in it anyway, and by virtue of this we can also see that 

- at least for the discipline of sociology if not for any lay interest in it - these 
things, ‘greetings,’ are of some central theoretical importance, though 
“Hello,” “Hello” looks like nothing that one would want to attend to very 
much. Their consideration does an enormous amount of work for us. And 
once we’re dealing with the fact that we’ve got sequencing here, and it’s 
regulated, we’re no longer in a position where linguistic investigations are 
usable. Because grammars don’t differentiate this way. 

With “Hello,” “Hello” and things like it - members of a class of paired 
activities such that if A uses one, B’s proper move is to use one also - we’re 
examining the sequential building blocks of conversation which are specially 
relevant in terms of their sequential character. Now, in that light, we can 
consider jokes. The following comes from a group therapy session; the 
members are teenagers. 2 

A: Hey wait I’ve got I’ve got a joke. What’s black and white and 

hides in caves? 

B: Alright I give up. What’s black and white and hides in caves? 

C: A newspaper 

A: No. Pregnant nuns. 


B: Whyn’t you run across the street and get me some more coffee? 

A: Why don’t you drop dead? 

D: Whyn’t you just run across-? 

A: What’s black and white thump black and white thump black 

and white thump? nun rolling down stairs. 

D: You know what a cute one is. You want to hear what a cute one is? 
What’s purple and goes bam bam bam bam A four-door plum. 

2 The fragment is taken from Sacks’ research notes and is slightly different from, 
and closer to the actual data than, the one in the transcribed lecture. 


Part I 

A: Terrific 

D\ I think it’s much better than about a black and white nun going 

A: No. that’s the new fad. Instead of having elephant jokes, now it’s nun 


B: Nothing. A nun. 

D: hmmmmm 

A: hehhhhh 

C: What’s black white and grey? Sister Mary elephant. 

A: hehhhhh 

B: Say whata nuns really do? they must have some function? 

C: r Nothing 

A: 1 They travel in pairs. 

B: Nothing nothing hehhh 

A: They travel in pairs. One nun makes sure the other nun don’t get 


B: You know what’s a ball. Whistle at ’em whistle at ’em when they 

walk down the street. 

D: You know they usually ah pray. 

A: Yeah. 

D: That’s about all. 

B: But the rest of the church does. Or they pray harder. 

A: They’re women who have devoted their lives- 

D: They marry God. 

A: -to God 

A: No. they’re women who have devoted their lives- 

B: They’re women who’ve had a bad love life and become nuns. 

A: -their lives their lives to uhm the devotion of the church. 

B: J.C. and the boys. 

D: We’re on an awfully bad God kick. 

A: OK let’s change the subject. 

The first thing that’s important about jokes is that to use one is something 
like buying a drink among a bunch of people: They come in rounds. And if 
some person tells a joke then every other person present has the right to tell 
a joke. So we can say about a joke when it’s used, that it’s a ‘first joke’ and 
that it will provide the occasion for each other person present to have a chance 
to talk, and to have a particular kind of chance to talk; that is, a chance to tell 
a joke. 

So what? So there are a variety of cases where you get more then two 
persons present, where exactly what the rights to talk are of the various 
persons present, may be quite obscure. For example, there may be large status 
differences or a variety of things like that, and how one goes about providing 
that each person can talk under such circumstances may be, then, a real 
question. For someone to use a joke on that occasion is then to give each other 
a set place to talk, and also to give him something to say. 

Lecture 12 


And in this group therapy setting, those facts are quite non-trivial. The 
persons are there for two hours, and that they keep talking is absolutely 
crucial, for the occasions of silence are extremely dangerous to all persons 
concerned. What happens when they’re silent is that various persons in the 
place now begin to look for a face that’s noticeable among themselves, and 
then pose for that person the problem of giving an account of why he is silent. 
Or, for example, if any given person is silent for any length of time, then their 
silence is a noticeable fact; something about which they can be questioned. 
Now, if they want to raise their personal problems they’re quite free to do so, 
but if they don’t want to, and given that silence is something noticeable, then 
it’s important to have some sorts of things that will permit everybody to talk, 
where they can talk without saying anything that can be ‘used against them.’ 
It also seems to be the case that any given body of talk, starting at any given 
place, will, if allowed to go on, end up dangerous. That is, it will end up on 
some topic which is perhaps too important to be talked about except under 
real feelings of relaxation. Things like God, death, sex, for example, which 
always come out, whatever topic is started. And persens are - and the people 
in this therapy group are - much given to watching when it is that a topic 
looks like it’s about to shift into something that from the group’s point of 
view is to be avoided - though any given person can talk about it 

What seems involved, then, is the development of things which permit 
talk to go on, and to go on in an ‘unaffiliated’ manner. Notice about jokes, 
that when jokes are told they’re things that are ‘going around;’ they’re quotes. 
So they’re unaffiliated remarks, and in that sense it’s hard to say about 
somebody that the fact that they told some particular joke has some special 
significance. They just heard it, and now they’re repeating it. 

Persons can then monitor the conversation, watching either for silence or 
for the approach of something dangerous, and start a bloc of talk by flicking 
in a joke, thereby giving each other person their chance to talk, and to talk 

I’ll just note here that there are other things which have the ‘unaffiliated’ 
character of jokes - that is, the speaker does not disclose his position by using 
it - but which don’t have that sequential character of going in rounds, which 
seem to get used in similar circumstances. At least these kids use them, and 
until I noticed that they were being used in similar circumstance to the jokes, I 
found them puzzling. What they do is, at points in the conversation when 
either nobody is talking or they haven’t talked for a while, insert slogans. 
They’ll just come out with a piece of an advertisement from the radio, or a 
jingle, or obvious quotations sarcastically said. Again, then, it seems that they 
go about monitoring when they ought to be talking or when silence seems to be 
present, and flick out these things which, again, have this unaffiliated character. 

Let me turn now to another sort of sequencing issue. I said about “Hello,” 
“Hello” that it’s a paired phenomenon, and that when the second one comes 
out, it may well be that the conversation is complete. And we can note that 
either one, or both, can be delivered in such a way as to provide for the size 


Part 1 

of the conversation. For example, by not slowing down at all as you pass 
somebody. You say “Hello,” the other says “Hello,” and there’s no indication 
by either party that there’s going to be any more to it than that. 

Now, some of the ways that conversations can begin, provide at least a 
slightly different set of sequential characteristics, although there are ways in 
which they’re related to paired beginnings. Perhaps the best way to introduce 
it is by just reporting what I did the last time I was trying to introduce this 
some material, so as to indicate right off that it wasn’t as trivial as it looks. 
Before I presented the phenomenon, I’d asked people in the class to write 
down the first lines of what they took to be ‘pickups.’ I got 60 first lines, of 
which just under 60 were questions. 

What I had wanted to be saying to them, and which they could see once 
they had those collections, was that a person who asks a question has a right 
to talk again after the question has been answered. So, with a question 
beginning, the conversation goes at least something like A-B-A. It can go on 
from there, or it can end like that. And that may be without regard to what 
the question consists of or what the answer consists of. Now, one way that the 
conversation can go on from there is that the person who asks the question can 
use his initial right to talk again, to ask another question, and the same right 
holds. So you can get indefinitely long chains, running, Q-A, Q-A, Q-A, etc. 
Eventually I’ll go over the special relevance for certain conversations of the 
‘chain’ possibility. It turns out to be extremely important. Whenever it 
happens to occur in a conversation - and it doesn’t necessarily have to occur 
in the beginning, but that point where somebody starts questioning - then the 
‘chaining’ possibility can be quite crucial to the way that the conversation 
goes. 2 

Now, of the sorts of questions that occur in first conversations, let’s begin 
by looking at those which have a close relationship to “Hello,” “Hello.” Note 
that the use of “Hello” is a regulated matter. It is the sort of thing which can 
be used to begin a conversation where two persons have some initial right to 
talk to each other, such that the fact that they happen to be physically 
co-present provides the occasion for the conversation. But, especially for 
things like pickups, the fact that the two persons are physically co-present is 
not sufficient grounds for them to begin talking, and “Hello” may be 
inappropriate. You can get conversations which go: 

A: Hello 

2 This, and the following materials, constitute a next run at some of the 
phenomena considered in lecture 7 of this Fall 1964-Spring 1965 series. These 
materials were not incorporated into that earlier lecture although such incorporation 
has been done with most of the Winter 1965 materials, because in this case it would 
introduce an anachronism. Specifically, in this second run we see the first reference to 
“the ‘chaining’ possibility,” which later crystallized as ‘the chaining rule’ (see Spring 
1966, lecture 2, p. 256 and lecture 2 (R), p. 264). Not incorporating these materials 
into the earlier lecture makes for some repetition, but the genesis of ‘the chaining rule’ 
seemed worth preserving. 

Lecture 12 

B: (No answer) 

A: Don’t you remember me? 


Where that involves proposing that there had been an initial right to use 

In the absence of some obvious warrant for the conversation to take place 
by virtue of two persons being copresent and nothing else, you get that sort 
of question which provides that although it doesn’t seem to be the case, there 
is indeed a warrant. There’s a whole range of things which tend to formulate 
a first conversation as a version of an nth. Things like "Don’t I know you 
from somewhere?” “Didn’t I see you at such-and-such a place?” "Didn’t you 
go to such-and-such a school?” “Aren’t you so-and-so?”, etc. All of these 
provide for the fact that it may be the case that we know each other, and if 
we do, then this conversation can take place as ‘a further conversation.’ In 
those cases, then, and more generally, we have a class of questions which 
provide an account for a conversation developing; that is, that this is not an 
initial conversation. 

There are a variety of other accounts which focus on different matters. For 
example, very frequently the first question will be a request. And the request 
will be such a thing as can be asked when any two persons are physically 
available to each other; you’re standing in a crowd waiting for a plane to 
arrive and someone asks, “When is the plane expected?” A variety of such 
informational matters can be offered. Note that for any one of these, to 
whatever the first answer is, another question can be constructed: 

A: When does the plane arrive? 

B: 7:15. 

A: Are you going to San Francisco also? 

It’s also to be noted that such standardized questions as "When does the 
plane arrive?” "What time is it?” etc., by virtue of the fact that they are 
standardized, provide for the relevance of the sequencing rules such that one 
knows when one of those questions is complete. Further, one knows what an 
answer to such a question looks like, so that the one who asked the question 
can know when the thing that stands as an answer will have finished, and thus 
provide that the other can talk again. This stands in contrast to discursive talk, 
where it may not be clear in that fashion when it is that somebody has 
finished. For persons who don’t know the discourse patterns of somebody 
they’re dealing with, the use of standardized objects to build the beginning of 
a conversation may be quite important. First of all, that you don’t wait too 
long after the other has stopped, where waiting too long might provide for 
their withdrawal altogether. Secondly, that you don’t interrupt, where of 
course one doesn’t want to be ‘rude’ to someone with whom one is making 
an effort to get acquainted. 

((Thereafter is a discussion of the question "Do you have a light?” See 
lecture 7, pp. 50-1.)) 

Lecture 13 
On Proverbs 

I’m going to talk about proverbs, trying to develop what’s interesting about 
them. I’ll begin by doing something I don’t normally do, which is to read you 
the way that proverbs are largely used by social scientists - because it’s quite 
relevant to the task they seem to have set themselves. The first quotation 
comes from page 3 of a book called The Study of Thinking by Jerome Bruner 
and some associates of his. 

That there is confusion remaining in the adult world about what 
constitutes an identity class is testified to by such diverse proverbs as 
‘plus ga change, plus la meme chose’ and the Hericlitan dictum that we 
never enter the same river twice. 

A very similar sort of remark comes from pages 1-2 (these always come in the 
first several pages of a book) of George Homans’ Social Behavior: its 
elementary forms. 

My subject is a familiar chaos. Nothing is more familiar to men than 
their ordinary, everyday social behavior; and should a sociologist make 
any generalization about it, he runs the risk that his readers will find him 
wrong at the first word and cut him off without a hearing. They have 
been at home with the evidence since childhood and have every right to 
an opinion. A physicist runs no such risk that the particles, whose social 
behavior in the atom he describes, will talk back. 

The sociologist’s only justification is that the subject, however 
familiar, remains an intellectual chaos. Every man has thought about it, 
and mankind through the centuries has embodied the more satisfactory 
of the generalizations in proverbs and maxims about social behavior, 
what it is and what it ought to be . . . What makes the subject of 
everyday social behavior a chaos is that each of these maxims and 
proverbs, while telling an important part of the truth, never tells it all, 
and nobody tries to put them together . . . every man makes his own 
generalizations about his own social experience, but uses them ad hoc 
within the range of situations to which each applies, dropping them as 
soon as their immediate relevance is at an end, and never asking how 
they are related to one another. Every one has, of course, every excuse 

A combination of Winter 1965, lecture (8), and lecture (9), pp. 1 and 8-12 
(transcriber unknown on these two) and Fall 1964, tape 13, side 1. 


Lecture 13 


for this shortcoming, if it be one. Social experience is apt to come at us 
too fast to leave us time to grasp it as a whole. 

I don’t intend to make any detailed comments about those remarks, except 
to note parenthetically that Homans’ procedure for starting a book is one of 
the most recurrent you’ll find, and in a way it’s enough to tell you about the 
kind of book you have here. He has to have an excuse to study the 
phenomenon he wants to study. That it happens is not a sufficient excuse. He 
has to show some problem. And he starts off with the supposition that persons 
think they know about the thing he wants to study, so he finds a way to show 
that they don’t. Now, to notice that is to notice that in the ordinary world, in 
everyday life, ‘engaging in inquiries’ is an accountable thing. Where, then, the 
work of sociologists remains constrained by that format. 

I want to focus on this sort of thing, available in both these quotes: It’s a 
very usual use of proverbs among academics, to refer to them as ‘propositions’ 
and to suppose then that it goes without saying that the corpus of proverbs 
is subjectable to the same kind of treatment as, for example, is scientific 
knowledge. They then build the basis for an inquiry - which has nothing to 
do with proverbs - by virtue of the fact that these propositions, when 
compared - without showing that they are actually compared in their use - 
are inconsistent. 

So one question is, are the collections of proverbs indeed a set of 
propositions in the sense that Homans proposes? Do you find any reason to 
suppose that that’s so? For example, here’s something from a newspaper: 

Premier Krushchev’s removal was viewed in Paris today as a serious 
blow to the improvement of relations between the Soviet Union and the 
Western powers. The primary reaction of government officials and 
diplomats was surprise. But they were also deeply concerned. “Better 
the devil you know,” one diplomat said, “than the one you do not 

Would somebody seriously say to this fellow, “What’s the evidence for that?” 
If it were a proposition in the first place, then the statement “How do you 
know?” which is used not only among scientists but is offered by Members, 
might occur. 

But one of the facts about proverbs is that they are ‘correct about 
something.’ That fact is especially important since some of them contain rules, 
and are invoked to govern various situations. Now, for many other kinds of 
rules, even in highly rule-controlled situations like the legal courts, if you 
invoke a rule by reference to precedent, the occasion of using it can provide 
the occasion for reconsidering that rule to see whether, not only in this instance 
but in general, it ought to obtain for anything. So that a rule introduced to 
govern a situation in a law case can be changed altogether. Which is to say, 
even a strict precedence system such as that, doesn’t have objects as powerful 
and as limitedly attackable as proverbs. 


Part I 

It’s in part that this is so that makes it quite irrelevant when a proverb was 
established - and in that sense, Bruner’s citing of “the Hericlitan dictum that 
we never enter the same river twice,” which is several thousand years old, is 
a correct use of proverbs - whereas it might be relevant in, say, a lawsuit, 
when a rule was established and what the circumstances were that generated 
it. So, for example, in current civil rights cases you occasionally find an 
extremely old statute introduced, where, then, the fact that it was introduced 
many years ago by reference to, say, the buying and selling of slaves, can be 
used to argue that it really isn’t worth anything now. For proverbs, such 
considerations are just quite irrelevant. Likewise, the organization or the 
society within which they were established is often almost as irrelevant. Where 
an American court would never think of using a Russian precedent, those 
Russian proverbs we come to know are treated as quite appropriate in 
ordinary discourse. For one, then, the object ‘proverb’ is enormously widely 
found, and, further, many proverbs are applicable quite across cultural 

Now, aside from the social scientists’ orientation to proverbs, anything I’ve 
ever looked at on that matter involves a list of proverbs and where they come 
from, their age, variations, etc. Nobody seems to deal with actual occasions of 
their use. And that’s because it is the folklorists, with their particular interest 
in proverbs, who have been collecting them. I want at least to make a start on 
considering proverbs in terms of occasions of actual use. 

There is a class of proverbs known as ‘proverbial phrases,’ one set of which 
I’ll focus on; things like “You’re stacking the deck,” “He’s hitting below the 
belt,” “You’re way out in left field,” etc. This set of proverbs comes from 
domains which have clear parameters. In baseball there are demarcated areas, 
so that “left field” is a locatable place within that domain. And “stacking the 
deck” is a locatable violation in playing cards. It seems to me that, for one, 
this sort of proverb may provide a clarification of the sense in which we might 
talk about ‘families of actions.’ And by ‘families’ I mean this kind of thing: 
Wittgenstein talks about ‘families of games’ and proposes, for example, that 
there is some intersection of rules between games in a same ’family.’ I once 
tried to see what such an intersection of rules might involve, with the 
following kind of trick. We took violations of rules in one game, which were 
not violations in another game, and began to use them just to see what might 
happen. So, for example, if you’re playing cards, it’s not proper to stand 
behind your opponent and look over his shoulder, but there’s no rule about 
it in chess. It’s nothing. So we did it in chess, and it did cause some kind of 
disturbances. There was a sense that somehow something must be wrong. 
And that perhaps had to do with the fact that it was a violation of a rule in 
some game. It seems, then, that there may be a sense in which rules in games 
can carry over into situations for which they haven’t been specified as rules. 

And for ‘families of actions’ we might talk about the following sort of 
situation. For some kind of activity there may be an event which is, in that 
activity, not regulated. It’s not even seen as an event; there’s no language for 
it, or if it’s pointed out descriptively, nothing much would be said about it. 

Lecture 13 


But if you invoke a rule from another activity - especially those which can be 
produced in proverbial form - then you find that you can come to see the 
event as indeed an ‘event,’ and as possibly illegitimate. So, for example, 
suppose we have a meeting, and there are various interests present at the 
meeting. The way in which persons arrange their seating may not be treated 
as something noticeable, and if pointed out would only be treated as, “So they 
happen to be sitting in some order, so what?” But with respect to playing 
cards, the way in which a deck is shuffled is something that’s regulated. And 
there are clear ways of violating that, which are observable, and which are 
talked of as “stacking the deck.” And when we propose at a meeting that 
“the Commies are stacking the deck,” then it’s clear what it is that’s going on, 
and it’s treated as something suspect. 

So apparently one uses proverbs of this character to make events noticeable, 
perhaps to make their ordered character noticeable, and then to formulate 
their ordered character by reference to their possible illegitimacy - where there 
is in the first instance no rule governing this or that particular event. We 
might loosely talk of this as a matter of ‘analogy,’ though it only becomes 
analogy once it’s proposed that some order is relevant in both circumstances. 

What we can come to see is that there can be a very limited set of 
paradigms or models, each of which may have demarcated areas of order 
which can operate generatively for an enormous range of further areas - 
without, however, it being necessary to produce a further set of rules at all, or 
to further demarcate the parameters of this other mass of events. Once one 
knows that “stacking the deck” is something that can be done with a set of 
cards, it can be seen that something like it’ is being done with a set of people. 
Or, for example, where “hitting below the belt” is something that is done in 
an exchange of blows, it can be seen as comparable to something that can 
happen in an exchange of remarks, and thus to operate in conversation and 
other sorts of exchange activities. One doesn’t have to construct a new 
language, but can retain the base source of this or that rule, as providing the 

I want to turn now to a consideration of proverbs as pieces of the language. 
In Archer Taylor’s classical book The Proverb, he mentions in passing one 
aspect of these things, which got me started on the line I’ll be discussing. 
Right at the beginning of the book he’s talking about defining proverbs: 

The definition of a proverb is too difficult to repay the undertaking; and 
should we fortunately combine in a single definition all the essential 
elements and give each the proper emphasis, we should not even then 
have a touchstone. An incommunicable quality tells us this sentence is 
proverbial and that one is not. Hence no definition will enable us to 
identify positively a sentence as proverbial. 

When I read the book a couple of years ago, what was important in that 
statement just never occurred to me. I’ll formulate it in what may sound a 
very queer way: If we take it with the linguists and the grammarians that the 


Part I 

sentence is constructed according to some rules, and that it is a kind of unit, 
then one question we might be led to ask is, what other ways, aside from 
linguistically, can we talk of a sentence as ‘a unit? What can be done with it? 
That is, what are the sorts of activities that persons can hope to accomplish 
within a single sentence? 

Such a question is obviously relevant to the analysis of social activities. 
Leaving proverbs aside for the moment, we could say that there would be at 
least a special kind of tension in a world where single sentences could do 
exceptional work. For instance, in the phone calls I collected at an emergency 
psychiatric hospital, you’d occasionally get the following. A man might report 
that he’s suicidal and offer as the immediate warrant that after 25 years of 
marriage his wife said, “Well, the fact is I never really loved you.” A 
statement like that has a dual consequentialness. It not only can bring 
something to a close, but it can provide that some series of events have been 
falsely seen; that is, it can erase them. 

And there is a whole set of things like that. Some religions are carried on 
that feature. The fact that you can know that at any last moment - if you get 
a last moment - you can always change your ways sufficiently to erase 
whatever it is you’ve done in the past, means that you do not have to order 
yours sins day by day. That is, no matter how your sins have added up, it’s 
possible at any last moment to change, and to get what anybody can get no 
matter what they’ve done earlier. Of course the fact that you may not get that 
last moment, itself sets up a variety of other kinds of considerations. But the 
sort of religious existentialism which focusses on every moment as a possible 
last moment, is itself only relevant given the notion of what can be done in 
any given moment. 

So the issue of what can be packed into some single unit is obviously 
interesting. And with respect to language, we can examine the variety of ways 
that the resources of a language have been explored in terms of how much can 
be packed into something while retaining such central features as, for 
example, transmissibility and reproducibility. Now, poetry and proverbs have 
taken a quite different tack from other explorations. It is said that poetry is 
very similar to mathematics, in the sense that you pack in knowledge with far 
more economy than prose would offer. And that is shown in elementary 
mathematics books when they ask you to “rewrite in English the following 
equation.” The same kind of test can be done for a poem. And it may be 
recalled that early Greek science, for example, did use poetical forms to write 
up the results. 

Now, one of the crucial things about proverbs is that they’re objects from 
an oral body of knowledge. They do indeed get written down, but their basic 
power and relevance seem to be as oral objects. There is a literature on oral 
traditions and how they’re preserved and used, and I’d like to suggest a most 
extraordinary book on this subject which is misleadingly titled if you’re 
looking for material on oral traditions, and that’s Preface to Plato by Eric 
Havelock. It is about Plato, but you could have no interest at all in Plato and 
learn an awful lot from it. Its basic concern is, what is Plato up to in his attack 

Lecture 13 


on Homer? Havelock argues that for the Greeks, Homer was an encyclopedia. 
His poems stored the enormous amount of relevant knowledge that the 
Greeks had to use, where the Greeks in this period did not, except in very 
exceptional circumstances, use writing. Homer’s poems were one of a variety 
of very powerful devices used to store that information. Plato’s concern was 
to break down that way of preserving knowledge because of a variety of 
things that bothered him about it. It’s not that Plato didn’t like poetry, in the 
modern sense of poetry, but that he was aware of the ways in which poetry 
is powerful - and the limits of that kind of power. 

I couldn’t begin to give an elaborate discussion on the relation of poetry 
and proverbs, but I can say a few things. Proverbs, like poetry, have a large 
use of metaphor, and they often have a kind of alliteration, rhyming, etc. 
There are more or less standard forms which are used, to which any given 
instance is fitted as far as I know; proverbs are formed as single sentences or 
phrases of sentences, they’re not longer than that. And, like poetry, their 
reproduction consists of the exact repetition of them - a poem and a proverb 
lose their character when they’re paraphrased. These sorts of features have 
some real advantages for maintaining a body of knowledge; we can say that 
they are constructed in mnemonically efficacious ways. That is to say, they’re 
very easily remembered and are thereby transmissible as ‘that very item’ and 
not in a paraphrase. Their stability, then, can be something independent from 
any occasion of use. 

Now for proverbs, I take it that one of the core features of their sense and 
of their use is that they are ‘atopical’ phenomena. So, for example, the sense 
and relevance of ‘a rolling stone gathers no moss’ is not found by reference to 
geological or botanical considerations. Some of the work of the neuropsychol- 
ogist Kurt Goldstein and his associates may be relevant here. One of the 
things they’ve found for children, brain-damaged persons, and sometimes 
among schizophrenics is that a kind of test devised by psychologists indicates 
that these people cannot handle proverbs - they don’t understand them, they 
don’t know how to use them. There are many protocols of persons presented 
with a proverb and asked to interpret it, and they produce long discussions 
about various features and behaviors of, for example, stones and moss. 

Goldstein proposes on the basis of those tests and other indicators, that 
there’s a big split between what he calls ‘abstract’ and what he calls ‘concrete’ 
thinking, and that persons who can’t use proverbs are persons involved in a 
failure of abstract thinking. But I take it that if you look at the protocols, the 
persons involved seem to be quite capable of dealing with proverbs 
‘abstractly.’ First of all, the proverbs themselves are quite abstract. ‘A rolling 
stone gathers no moss’ doesn’t contain any reference to a particular rolling 
stone, a particular kind of moss, etc., etc. And these people who are not able 
to deal with proverbs properly are nevertheless talking abstractly. They’ll talk 
about “a stone” and how it might roll, and say this or that about moss, 
without any insistence that it has to be some particular stone. 

That seems to me to stand in contrast to ‘concrete’ thinking. So, let’s say, 
when you talk about some kinds of schizophrenics being “enmeshed in the 


Part l 

concrete,” you might be pointing to certain kinds of strange things they can 
do that nobody else can do. For example, one day I was sitting having lunch 
with a friend of mine and she said, ‘‘What day is it today?” I told her what 
day it was, and she said “Last year on this day I had the following for 
lunch . . . two years ago this day I had the following for lunch ...” and she 
went through ten years, just spieling out the details of her menus. That, I take 
it, is a pretty clear example of concrete thinking. But the troubles in dealing 
with proverbs were not of that sort. Again, if the proverb contained the item 
‘cat,’ then people would talk about cats and use the plural term ‘cats.’ They 
did not start talking about ‘my cat.’ 

This suggests that we may not be dealing with inability to do ‘abstract’ 
thinking, but an inability to do ‘atopical’ thinking. Where, then, proverbs can 
be seen to constitute a very clear example of whole collections of pieces of 
knowledge that are organized atopically. And I take it that it may be this 
feature that I’m calling ‘atopical organization,’ that Homans proposes as a 
possible ‘shortcoming,’ in the remarks I quoted earlier: 

What makes the subject of everyday social behavior a chaos is that each 
of these maxims and proverbs, while telling an important part of the 
truth, never tells it all, and nobody tries to put them together . . . but 
uses them ad hoc within the range of situations to which each applies, 
dropping them as soon as their immediate relevance is at an end, and 
never asking how they are related to one another. 

But there are some obvious virtues to having a body of knowledge organized 
in an atopical fashion. You get a piece of knowledge, like ‘a rolling stone 
gathers no moss,’ which is in the first instance correct about something. If you 
paraphrase it into some particular domain, like “a man who doesn’t settle 
down doesn’t gather possessions,” then it may not have the same kind of 
correctness; it may be questionable. And as I mentioned earlier, one of the 
most striking things about proverbs is that while on any occasion of use they 
may be, for example, inappropriate, people do not propose about them, “Is 
that so?” “What is your evidence for that?” If there is a question about them, 
it is in terms of, is it appropriate to apply that proverb to this person, activity, 
etc.? That is, proverbs are in the first place correct. And that can be 
accomplished by formulating a proverb from a domain within which it is 
correct, and having it always be used elsewhere. In that way, instead of 
constantly revising a body of knowledge by reference to the discovery that it’s 
not correct here, now, for this, you maintain a stable body of knowledge and 
control the domain of its use. 

I’ve already mentioned the feature of ‘single-sentence packing’ for prov- 
erbs. Let me offer a few more remarks. It’s the case that both maxims and 
descriptions can be produced as proverbs, formulable as single sentences. “A 
woman’s place is in the home” is an instance of the former, and for the latter, 
they are frequently those ‘proverbial phrases,’ “stacking the deck,” “hanging 
by a thread,” “barking up the wrong tree,” etc. Both types of knowledge, 

Lecture 13 


then, can be had and used via the single sentence, and thus it’s not necessary 
to have some combination of sentences so as to minimally understand and 
transmit such information. Having that, you have a setup designed to permit 
you to learn new members of that class of information much more quickly 
than might be otherwise possible. That is to say, at that point when children 
can make single sentences, their task is now to see what it is that can be 
packed into that form they’ve learned. And that provides that they will have 
solved the theoretical problems of being able to know norms, or know 
knowledge. They would not then be faced with such a task as learning which 
set of sentences can constitute a piece of knowledge. Rather, once the single 
sentence is gotten, a whole series of things can be fitted to it, instead of having 
to deal with various combinations, which might set up quite different sorts of 

In this regard, an extremely important thing about proverbs is that they 
have the character of being potentially descriptive or relevant. Persons learn 
them and have them available for use. They don’t, that is, learn them on the 
occasion of their appropriate use. We could imagine that humans were built 
such that their language was alike to that of other animals, and to the way in 
which young children very heavily use language; either as narrative comment 
on what they’re doing, “Now I’m putting on my shoe,” or only uttering 
something on the occasion that it’s appropriate - though it may be wrong, 
they could say “That’s mommy” and it’s not mommy; that is, the domain of 
application would be correct or incorrect, but the thing is not uttered apart 
from some possibly appropriate occasion. But you can have these potential 
descriptions and see them as ‘correct for something,’ where what it would be 
correct for remains to be seen. 

And if you watch kids learning language, then learning that sort of thing 
seems to be an important part of it. Observations have been made of children 
talking to themselves before they fall asleep. And it’s been found that they’re 
not coming out with a bunch of random noises, but with a training procedure. 
For example, they’ve been observed doing exercises in phonetics, combining 
and assembling phonemes. I don’t think it’s been pointed out that once they 
have words, they play with combinations of terms that involve building 
possible descriptions. We could say it would be strictly a matter of learning 
the linguistic features of the language - learning the language in this very 
technical sense - that kids begin to use, in close relation to each other, the 
paired antonyms (things like in-out, up-down, etc.); that is, to produce these 
paired items as pairs. However, if they not only did that, but did - as I think 
they do - use them in proper sequence, then we’ve got them learning 
something more. They’re learning potential descriptions. That is, they don’t 
say, for example, “He fixed it, he broke it,” but “He broke it, he fixed it.” 
And for such things as in-out, which in some cases have proper sequences and 
you don’t turn them around, then they do them in proper sequence, “Kitty 
got into the box, he got out.” And if they do turn them around, they do it 
properly: “Kitty got out of the box, he got in again.” They’re learning, then, 
not only the relation of certain types of words, but that strings of words can 


Part I 

assemble potential descriptions, only in certain arrangements. And that is to 
say, they’re learning that there is a correct way of assembling potential 
descriptions, apart from the particular occasions of their use. 

Now, the phenomenon of ‘potential description’ has a variety of special 
consequences. One of them is that it sets up the possibility of a logic which 
involves examining statements apart from an occasion of use - and by ‘an 
occasion of use’ I mean an application to some actual situation, not, for 
example, an attempt to make a proof. Another is something which I had to 
be brought to see as noticeable, and that is the possibility of literature; stuff 
that is about nothing in particular, in which an author isn’t talking about any 
actual set of events. It occurred to me that the phenomenon of ‘the possibility 
of literature’ is something noticeable while I was looking at psychiatric reports 
about delusions that patients have. In these reports, the writers take it that the 
delusions are understandable, though what makes them ‘delusions’ is that 
they couldn’t possibly describe something. Literature has a similar character; 
it’s composed of possible potential descriptions and possible potential rules, 
and a reader can look at assembled strings of language and decide that it’s 
‘realistic’ or ‘not realistic,’ compelling or not compelling. And in that sense, 
the possibility of literature and the possibility of logic are very, very closely 

These sorts of considerations may have a bearing on a classical controversy, 
mainly within linguistics, which concerns the question: Are grammar and 
meaning separable? It may be that “Are they separable?” is not the problem, 
but that if ‘meaning’ is, in part, reference to something, and grammar is 
understandable apart from reference to anything - that is, formally correct in 
some way - then it’s not simply a technological linguistic question, whether 
it’s so that they’re separable, but that it’s an essential fact of language that 
they are separate. 

Lecture 14 

The Inference-Making Machine 

(1) A: Yeah, then what happened? 

(2) B: Okay, in the meantime she {wife of B} says, “Don’t ask the child 

nothing.” Well, she stepped between me and the child, and I got 
up to walk out the door. When she stepped between me and the 
child, I went to move her out of the way. And then about that time 
her sister had called the police. I don’t know how she . . . what she 

(3) A: Didn’t you smack her one? 

(4) B: No. 

(5) A: You’re not telling me the story, Mr B. 

(6) B: Well, you see when you say smack you mean hit. 

(7) A: Yeah, you shoved her. Is that it? 

(8) B: Yeah, I shoved her. 

One of the basic things I want to be able to give you is an aesthetic for social 
life. By that I mean in part that we should have some sense of where it is deep, 
and be able to see, and to pose, problems. I’ll try to do somewhat more than 
that. I’ll also try to develop a variety of notions of what kind of business 
sociology is, what its problems look like, what the form of the solutions to 
those problems are, and perhaps to some extent, some of those solutions. 

The kind of phenomena we are dealing with are always transcriptions of 
actual occurrences, in their actual sequence. And I take it our business is to try 
to construct the machinery that would produce those occurrences. That is, we 
find and name some objects, and find and name some rules for using those 
objects, where the rules for using those objects will produce those objects. And 
we also consider conversation per se, looking at the rules for sequencing in 

The quotation I started off with comes at about the fifth interchange into 
a first telephone conversation, where A is a staff member of a social agency 
that B has called. B was told to call this agency because of some marital 
troubles he’s having. A doesn’t know anything about B’s marital troubles, 
except what B tells him. So apart from the four or five previous interchanges, 

A combination of Spring 1965, lectures 1 and 2(transcriber unknown), with 
materials from Fall 1964, tape 12, side 2, pp. 1-6. Only three of the Spring 1965 
lectures are extant. Lecture 3 does not appear in this volume. The large part of its 
materials are covered in lecture 1, and a bit of it has been incorporated into that 



Part I 

these persons do not know each other, they’ve never met. And - though it’s 
not exactly correct to say it this way - we find that “nonetheless,” given 
essentially B’s statement number (2), A is able to have some notion of what 
it is that has happened, in a story that he has only heard part of. Without 
knowing B at all, hearing “When she stepped between me and the child, I 
went to move her out of the way. And then about that time her sister had 
called the police,” A can say, “Didn’t you smack her one?” 

And there is more to it than that. Because it would be one thing if A had 
some expectation about what the sequences of events were that brought on 
the coming of the police, where A would use that expectation to make a 
guess, but if it turns out that he’s told it’s not so, then so far as he knows, it’s 
not so. But apparently A has a stronger grasp of the situation already, in the 
sense that it’s not merely that he can have a guess, but that he takes it 
that - perhaps without regard to what B tells him - his guess is correct. 

And we can notice that the fact that A has a guess here is not anything very 
extraordinary for B. He doesn’t say, as a person sometimes says, “What 
makes you say that?” He says “No.” And when it’s proposed that his “No” 
is not correct, he doesn’t say, “Look, I’m telling the story. How the hell would 
you know?” (Parenthetically, when I say “He doesn’t say . . .,” that is a 
rhetorical device on my part for loosening things up. I don’t mean it in a 
serious sense. I am not making a statement that is intendedly descriptive. To 
propose seriously, descriptively, ‘B doesn’t say X’ is another order of 
proposition.) But B takes it that A does know, and B corrects himself. 

Furthermore, A is able to see as well, that B “isn’t telling the story.” And 
I want to briefly focus on that kind of thing: Seeing lies. I want mainly to 
focus on it because it seems that adult intuition may be misleading in trying 
to see what the problem is that seeing lies poses. In one of the early classics 
of psychoanalysis, a paper entitled ‘On the origin of the influencing machine 
in schizophrenia,’ which was published in English in the Psychoanalytic 
Quarterly, volume 2, in 1933, Victor Tausk reports on one of his patients. 
The patient was a young schizophrenic girl, and one of her symptoms he 
found in the following way. He often questioned her, and one time when he 
was questioning her she started laughing at him. He asked, “What are you 
laughing at?” She said, “Why do you always ask me questions? After all, you 
know what I’m thinking.” 

And Tausk worked at this symptom, that schizophrenics think other 
persons know their thoughts. The problem had been posed: How is it that 
schizophrenics come to think that others know their thoughts? And he tries to 
solve this problem. Now, Freud’s comments upon the presentation of the 
paper are included in the journal publication. He says, “That’s not the 
problem at all. After all, when you learn at least your first language, you learn 
it from your parents, from adults. And children must take it that adults, 
giving them the concepts, know how they’re being used; know how the child 
is using them. So the problem is not how is it that people come to think that 
others know their thoughts, but how is it that people come to think so deeply 
that others don’t know their thoughts?” Then, in a characteristic type of 

Lecture 14 


observation, Freud says that the crucial event is the first successful lie. That 
event must be traumatic. The kid must have to say, “My God, they don’t 
know what’s going on!” 

Whether that’s so or not, I certainly don’t know. But intuitively I take it 
that most adults don’t see things that way, and that it’s certainly not an 
inconceivable formulation. And the interchange we re looking at has a form 
which is very characteristic between adults and children; that is, the mother 
says to the child, “What were you doing?” the child gives some answer, the 
mother - who wasn’t there - says, “No you weren’t,” and the child then cor- 
rects itself. And again, it isn’t the case that persons proposing that one is lying, 
or “not telling the story;” are treated as doing something quite extraordinary. 

Now let’s try to begin to consider what our task is if we’re going to build 
a machine that could in the first place produce this conclusion, “Didn’t you 
smack her one?,” with this piece of information, “When she stepped between 
me and the child, I went to move her out of the way. And then about that 
time her sister had called the police.” 

A first rule of procedure in doing analysis, a rule that you absolutely must 
use or you can’t do the work, is this: In setting up what it is that seems to have 
happened, preparatory to solving the problem, do not let your notion of what 
could conceivably happen decide for you what must have happened. So, for 
example, when we get this kind of conclusion drawn at statement (3), you 
might say to yourself, how in the world could anybody think so fast? Because 
this just comes off. There is no pause more than three seconds between (2) and 
(3). You’re going to say, “How? People aren’t that smart.” And therefore 
what happened must be something very simple; something which will require 
only a simple solution. 

And that leads to our second rule. There is no necessary fit between the 
complexity or simplicity of the apparatus you need to construct some object, 
and the face-value complexity or simplicity of the object. These are things 
which you have to come to terms with, given the fact that this has indeed 
occurred. And insofar as people are doing lay affairs, they walk around with 
the notion that if somebody does something pretty simply, pretty quickly, or 
pretty routinely, then it must not be much of a problem to explain what 
they’ve done. There is no reason to suppose that is so. I’ll give an analogical 
observation. In a recent review of a book attempting to describe the 
production of sentences in the English language - a grammar, in short - the 
reviewer observes that the grammar, though it’s not bad, is not terribly 
successful, and it remains a fact that those sentences which any six-year old is 
able to produce routinely, have not yet been adequately described by some 
persons who are obviously enormously brilliant scientists. Of course the 
activities that molecules are able to engage in quickly, routinely, have not been 
described by enormously brilliant scientists. So don’t worry about the brains 
that these persons couldn’t have but which the objects seem to require. Our 
task is, in this sense, to build their brains. 

Now, what features do we need to build into this machine? the first thing 
of course, is that it is an ‘inference-making machine.’ That is to say, it can deal 


Part I 

with and categorize and make statements about an event it has not seen. And 
the first thing about the sort of events it can handle is that they can be 
sequential events. In the interchange we’re looking at, if we keep things very 
simple what we see is that if we have (a), (b) . . . (d), then if we suppose that 
A is a user of the machine we’re going to have to build, he can find what (c) 
is. He is not merely in a position to make a guess about (c), but there are 
stronger features to this machine. As I said earlier, it’s not that he guesses 
something, is told “No” and says “Okay.” For him, there is some fit 
operating between (a), (b), (c), and (d) such that the fact that he’s told that 
what he proposes as (c) has not occurred does not constrain what he can 
continue to insist on; that it was that (c) which occurred. And you can just 
quickly contrast that to the situation of a riddle. I ask you a riddle and invite 
you to give an answer. You give an answer and I say “No, the answer is . . .” 
something else. And by definition that’s the answer. Where people do riddles 
you don’t get an insistence, “No, I’m right.” The inference-making machine 
we are building can handle riddles, but riddles set up a simpler task than 
those which this machine is capable of handling. 

Now let’s begin to examine, in a rather informal way, how it is that the 
machine-user, A, seems to make conclusion (3), “Didn’t you smack her one?” 
and then (5), “You’re not telling me the story, Mr B.” And here we can use 
that information which we have as members of the same society that these 
two people are in. What we have is roughly something like this: A knows that 
the scene is ‘a family problem.’ So (a) is the family quarrel, (b) is the guy 
moving to the door . . . (d) is the police coming. And (c) is the grounds for 
the police to have come. That is, apparently on some piece of information the 
police have come, and that piece of information is the thing that A has 
guessed at. A apparently knows, then, what good grounds are for the police 
to be called to a scene. And he’s able to use those good grounds, first to make 
a guess, and then to assess the correctness of the answer to that guess. 

We can note as well that it’s not simply the case that A and B don’t know 
each other, but we have a set of other persons who are being talked about, and 
A is listening to this. Whatever A knows about B, he certainly knows less 
about these others. A knows essentially only the set of terms that B uses to 
name them; that is, that there is something called a sister, something called 
a wife, something called a child. It seems to me that the information that is 
being used by A is held in terms of collections of these categories. For one, you 
can easily enough come to see that for any population of persons present there 
are available alternative sets of categories that can be used on them. That then 
poses for us an utterly central task in our descriptions; to have some way of 
providing which set of categories operate in some scene - in the reporting of 
that scene or in its treatment as it is occurring. 

To get a sense of the way in which the inferences that can be made from 
a story are geared to these categories, we could try, for example, using 
different categories. What if it were, not “her sister,” but ‘a neighbor’ who 
had called the police? A possible inference in that case would be that the 
grounds for calling the police had something to do with ‘creating a 

Lecture 14 


disturbance;’ crying child, husband and wife yelling at each other. Or, for 
example, just shuffle the one category around a little bit. Would the same 
inference be made if it was ‘my’ sister, not “her” sister who had called the 
police? The rules with respect to who owes what to whom, and who takes care 
of whom may be so formulated that those things matter a great deal. The 
inference in this case might then be, not that the husband had produced some 
activity which served as good grounds for calling the police, but that the wife 
had done so. 

And that is extremely important because it is an awesome machine if one 
needs to know only that it is “my wife” and “her sister.” And you can do this 
because that holds for every like unit in society, such that you don’t need to 
ask for example, “Well tell me some more about your wife’s sister, is she 
elderly? Is she prone to hysterics?” which is something that would be 
absolutely essential in psychology. But if something like what I’ve been saying 
is so, then: It is not merely that the notion that you need to know a great deal 
about somebody before you can say this or that about them may be a lot of 
nonsense, but the way that society goes about building people makes a 
nonsense of such a notion. That is: A task of socialization is to produce 
somebody who so behaves that those categories are enough to know 
something about him. 

One of the things we always want to be watching for is to see how simple 
or how complex this animal is. In this regard I’ll raise a question but not 
attempt to answer it here. There is what we can call an ‘order of depth’ in 
dealing with various kinds of occurrences. So, for example, we can say about 
A here that there can be a list of good grounds for police to be called to a 
scene, and that list is built into his brain as a mature member of the society, 
and he can, when told of some scene where the police come, now throw out 
an item from that list as a guess, and perhaps furthermore insist on it being 
correct. That would be one order of depth. But for the issue of how simple 
or complex this animal is, we can notice that, for example, for dealing with 
such a situation as ‘the police arriving on the scene,’ we need a machine; a set 
of rules, that is, and a set of objects that those rules handle. We can then ask, 
is that machine altogether distinct, such that if something else happens, we 
need another machine for making another kind of inference, etc., etc.? Or will 
it be the case that this use of an inference from the ‘police’ situation will be 
absolutely similar to the way in which some other event is found? Then 
police’ just becomes one category of a general machine that handles a whole 
set of things. If that is so, then the task of describing how this animal operates 
will be tremendously simplified. And the work of the animal itself is, of 
course, tremendously simplified. 

What we’re working with, for now, is that apparently A knows what the 
good grounds are for the police to come, and uses those good grounds to 
monitor what B tells him, and makes a guess and then challenges B’s assertion 
that he’s wrong, I want show how this stuff can cut; how it can be interesting. 
For example, it’s clearly the case - A knows it and B knows it - that the 
police don’t only come for the good grounds that police come for. Cranks call 


Part I 

up the police all the time. And police answer calls that turn out not to be 
based on good grounds. But apparently that fact has a special status. The 
proper ways that the police come to some house are available to Members in 
general, and it is those which may be used, for example, to decide what 
happened at that house. And if you see somebody being led away by the 
police, you may quite naturally feel that you know what’s up with him, at 
least generically; that is, that he’s ‘done something wrong.’ 

Now that suggests two things: First, the fact that some procedure which 
has a correct way of getting done, gets done correctly - independent of any 
issue of police answering calls that are not based on good grounds, or any issue 
of people getting harassed or misunderstood — may be quite crucial in 
permitting persons to find the sense of an event which happened by reference 
to those procedures. And second - and here we’re going to pose one of the 
central dilemmas of Western civilization - a person who stands in the position 
of having some procedure which has a correct basis for use applied to them, 
stands in the position of having that procedure presumptively correctly 
applied. And we can give that problem a name: Job’s Problem. 

Job, the rich, good man, had lost all his wealth, his children, all his 
possessions. His friends come to him, and there are series of long discussions. 
What his friends propose is, look, you take it that God punishes the wicked 
and rewards the good. We take it the same way. Your situation is 
understandable only if you’re guilty. So confess. And for Job the question is, 
“I don’t know that I’m guilty. I’m convinced that I’m not guilty. But then 
how could this have happened to me?” 

A central dilemma, then, is that some procedure which has a proper way 
of operating, may not so operate. Kafka is dealing with the same kind of 
issue. And Mr B here is in the same boat. What A suspects may indeed have 
happened. Or B may have just felt . . . what can he say? After all, A knows 
how those things happen, and how is B going to insist that it’s not so? That 
is, a problem that people in a range of circumstances can be faced with is: Is 
there a way available to provide from some event to have happened, apart 
from the normal and proper way these things happen? We are talking about 
something quite general here. Talking about it by reference to the police 
doesn’t mean we are talking about a procedure set up by and for the police. 
Rather, it may be the case that in order for the police to operate successfully, 
they have to be able to produce their activities such that these ways that 
Members have of looking at activities can be applied to the police. Likewise, 
when I talk about Job, I’m not trying to deal with literary criticism. Job’s 
problem focusses on a central problem in the way that persons go about 
orienting to the occurrence of events, and that is, that it is somehow extremely 
important that the inferences they do make can be taken as correct, and 
thereby that those persons who produce those activities which are described by 
these sequences so behave as to provide for the fact that these sequences do 
describe them. 

In that regard it is interesting to note that the phenomenon of presump- 
tively correct descriptions, and behavior produced to fit those descriptions, can 

Lecture 14 


be found by reference to illegitimate as well as legitimate activities. So, for 
example, in his book Order and Rebellion in Tribal Africa, in the chapter ‘The 
reasonable man in Barotse law,’ Gluckman offers us the ‘reasonable wrong- 

The last case suggests that the Barotse have a picture not only of 
reasonable and customary right ways of behaving, but also a picture of 
the reasonable wrongdoer - the reasonable thief, adulterer, slanderer, 
and so forth. 

By this paradox - the reasonable wrongdoer - I sum up the fact that 
wrongdoers in any society also behave in customary ways which are 
socially stereotyped. There is the ‘criminal slouch’ as against the 
‘scholarly stoop,’ the spiv’s clothes and hairstyle, the whole manner of 
loitering with intent to commit a criminal action. 

When there is only circumstantial evidence, these sorts of actions 
build up before the judges until they conclude that . . . the total picture 
is that of a reasonable wrongdoer; as we say, a person guilty beyond 
reasonable doubt. 

And I have heard cases which indicate that these customary ways of 
doing wrong in fact influence adulterers and thieves, so that they give 
themselves away in circumstances in which they could have acted so as 
to cover up their misdeeds. 

Our task, then, is not simply to be building a machine which can make 
inferences, and make them in the strong way that I’ve proposed. A problem 
for a sociology interested in describing socialization will consist in large part 
of how it is that a human gets built who will produce his activities such that 
they’re graspable in this way. That is to say, how it is that he’ll behave such 
that these sequencing machines can be used to find out what he’s up to. Under 
that notion, then, we would propose that at least one core focus for trouble 
would be persons who are so socialized that they don’t permit these 
sequencing machines to be used on them. And that is one way that 
‘psychopathic personalities’ behave. If you read Cleckley’s book The Mask of 
Sanity, the psychopathic personality is reported to be that person who, at any 
given point in their behavior, you never know what’s going to happen next. 
You’re never able to say “Here is an nth point in this sequence, and now X, 
Y, and Z will come.” And they are taken to be about as painful a person as 
you can have around you. 

Now, what I have been proposing could be restated as follows: For 
Members, activities are observables. They see activities. They see persons 
doing intimacy, they see persons lying, etc. (It has been wrongly proposed that 
people do not see, for example, ‘my mother,’ but what they ‘really see’ is 
light, dark, shadows, an object in the distance, etc.) And that poses for us the 
task of being behaviorists in this sense: Finding how it is that people can 
produce sets of actions that provide that others can see such things. 


Part I 

Earlier I observed that one way that persons go about seeing activities is by 
reference to some procedures which they take it properly occur as the activities 
occur. That is, it can be seen that Mr B “smacked” his wife by virtue of the 
fact that the police came, by virtue of the fact that the police come when 
people smack their wives. While in many of its aspects the use of procedures 
to find the sense of some set of observables - that is, the phenomenon I’m 
calling Job’s Problem - is utterly central, this fact of observability of actions 
is much more generic. 

We can be led, then, to investigate, for example, how it is that persons 
learn that by virtue of their appearances, the activities they have gone through 
are observable. Where, again, this observability is not specific to each activity, 
but is learned as a general phenomenon. And we could suppose that the 
following sort of report might be the sort of thing we could use to find out 
how this learning takes place, and to see where those things may or may not 
be difficult. I will take this utterly mundane report and suggest some of its 
relevance. It is a quotation from One Boy’s Day by Barker and Wright (195 1). 
What they did was to have a bunch of people follow a kid around all day, 
writing down as best they could, everything that he did. They worked in 
half-hour shifts, and they compiled, then, a record of his day. 

7:20 Raymond got up from his chair. He went directly out of the 
kitchen and into the bathroom. 

Coming from the bathroom, he returned to the kitchen. His 
mother asked pleasantly, “Did you wash your teeth?” Mr Birch 
looked at him and laughed saying, “My gosh, son, you have 
tooth powder all over your cheeks.” Then both parents laughed 

Raymond turned instantly and went straight to the bathroom. 

He smiled as though he were not upset by his parents’ comments. 

He stayed in the bathroom just a few seconds. 

He came back rubbing his face with his hands. The tooth 
powder was no longer visible. 

Presumably Raymond can learn through things like this, that his parents 
can tell that he washed his teeth by virtue of the appearance on his face of 
tooth powder. And that fact sets up the phenomeon which I’ll call generically, 
‘subversion.’ With this example in hand one can think quite rapidly of the 
way that children can learn subversion, having learned that the procedure is 
applied to them. For example, they will wash those aspects of their body 
which are at a glance observable, so when they appear, it seems “Oh, you 
washed.” And the fact that that adaptation goes on, provides, then, the sense 
in which they orient to this way that their activities are grasped. 

Now, in watching these obviously trivial things that children might do, it 
is conceivable that one may have analyzed matters that are treated as being of 
larger moment. So, for example, the quotation I just gave, in which 
Raymond’s having brushed his teeth was seen in his face covered with some 

Lecture 14 


white powder, is in its character extremely central, and deserves the name 
‘generic.’ Simply enough, the first human event in Judaic-Christian mythol- 
ogy consists of man’s discovery that his moral character is observable. We 
might call it Adam’s Problem. 

(Genesis, 3.6-12, King James Version) 

And when the woman saw that the tree was good for food, and that it 
was pleasant to the eyes, and a tree to be desired to make wise, she took 
of the fruit thereof, and did eat, and gave also unto her husband with 
her; and he did eat. And the eyes of them both were opened, and they 
knew that they were naked; and they sewed fig leaves together, and 
made themselves aprons. 

And they heard the voice of the Lord God walking in the garden in 
the cool of the day; and Adam and his wife hid themselves from the 
presence of the Lord God amongst the trees of the garden. 

And the Lord God called unto Adam, and said unto him, Where art 

And he said, I heard Thy voice in the garden, and I was afraid, 
because I was naked; and I hid myself. 

And He said, who told thee that thou was naked? Hast thou eaten of 
the tree whereof I commanded thee that thou should not eat? 

And the man said, the woman whom Thou gavest to be with me, she 
gave me of the tree, and I did eat. 

And Adam learns, like Raymond learns, that one has to come to terms with 
the fact that from one’s appearances the activities that one has engaged in are 

Let me mention one further thing, which is again relevant to the 
conversation we’re examining. The matter I’ll be dealing with is one which 
you might not, given the way one goes about considering scientific materials, 
formulate in the way I’m going to. When we think about facts, insofar as we 
are thinking of scientific facts, we tend to pose problems in the following way: 
If it’s the case that something has occurred, then our problem is to explain it. 
Now, with such things as lies, untruths, confabulations - the possibility of 
which persons are often attending to - we’ve got to notice that something like 
a reverse procedure is very much used. The reverse procedure consists of the 
following. In deciding among possible competing facts, one may decide that 
that fact occurred which has an explanation, and that fact that hasn’t an 
explanation did not occur. Here is a beautiful instance of this way of 
proceeding. It comes from an arbitration case involving a company which 
discharged some of its employees for causing a disturbance. Here’s what the 
arbitrator writes: 

(Shulman Decision A-70 Ford-UAW Arbitration) 


Part 1 

The story of the other discharged employees approaches the bizarre. 
Twelve of them testified before me. Each of them claims to be a 
completely innocent bystander, wholely at a loss to understand why he 
was picked up for discharge. None of them admits being part of the 
crowd in any of the demonstrations. None of them admits even the 
normal curiosity of an innocent bystander. Each claims that he knew 
very little about the cause of the stoppages, and cared even less after 
learning the cause. Each claims that when the lights went out or the line 
stopped he asked his foreman what to do and on being told to stay or 
go home as he pleased but that his time stopped in any event, he left for 
home. One of the men, a lively young boxer, asserts that after seeing the 
crowd and the excitement he calmly repaired to a warm, comfortable 
spot and went to sleep. He did this, he asserts, on two of the three days 
(being absent on the third), and slept the peaceful sleep of the just, until 
the excitement completely quieted down. All, it seems were veritable 
angels, above and beyond contagion by the excitement in the depart- 

Now there unquestionably were serious stoppages in Department 84 
on November 5th, 6th, and 8th. There were vociferous and angry men 
milling around and demanding action. Who were the incensed men 
who did take part? Who were the angry men who were so difficult to get 
back to work and were so incensed that, the union claims, they turned 
against their own committeemen and even assaulted two of them? How, 
indeed, were these fourteen chosen? 

The union advanced no explanation. There is no suggestion that these 
men were chosen by lot. And even such a method would normally be 
expected to catch some of the guilty. And there is no basis whatever in 
the evidence to suppose that the men were selected because of any 
personal animosity against them, with the slight possible exception of 
one man. Nor are the men generally regarded as troublemakers, of 
whom the company would be glad to be rid. 

The company’s explanation is simple and without any contradiction 
other than the incredible stories related by the men themselves. The 
labor relations conciliator, with the help of his assistant, took the names 
or the badge numbers of the most active men in the crowds that 
demonstrated in his office. This accounts for twelve of the fourteen. 

. . . Under these circumstances I cannot give credence to the men’s 
protestations of innocence. 

In this case the two competing facts are that a group of employees are 
innocent, were wrongfully discharged, and that they are guilty, discharged on 
good grounds. If they are innocent, there is “no explanation” for their being 
selected to be fired. If they are guilty, then the “explanation is simple and 
without any contradiction.” The finding is that they are guilty. Here is 
another instance of that sort of procedure, taken from a coroner’s report. The 
competing facts in this case are death by suicide or accident. 

Lecture 14 


While Mrs S.’s drinking during the past twenty years may be 
symptomatic of problems, there was nothing in the history to indicate 
any sudden change in her life pattern, or any unfortunate or untimely 
occurrence in her life. And therefore there seems to be no reason why she 
should have chosen this particular time to end her life. 

That is to say, since there is no reason for her to have ended her life, she didn’t 
end her life. 

This gets very, very subtle and curious at times, where you find the 
following kind of situation. One recurrently problematic kind of death 
involves alcohol and pills in some combination, where alcohol and pills in 
combination are extremely deadly. That is to say, given some amount of 
alcohol, far less pills are necessary to kill somebody than without alcohol. And 
where persons die of alcohol and pills, with a few pills, then you get 
something like the following argument proposed: It’s not suicide, it’s an 
accident, because if they wanted to kill themselves they would have taken all 
the pills they had, and they didn’t take them all. To which you might say, 
well, but they did die, and maybe what they did was to kill themselves in a 
perfectly efficient way. 

One of the nicest cases I have is from the autobiography of an ex-mental 
patient. It’s a very long description so I’ll give it in condensed form. He’s just 
committed himself, he’s been there maybe a day or two and now he wants to 
inform his family, but nobody will let him out of the ward to make a phone 
call. He finally asks a doctor, who tells him “Wait here five minutes, I’ll look 
for your file. If I find it you can be let out to make a phone call.” So he waits. 
He stands there waiting for five hours. At some point a nurse comes by and 
tells him to move away because he’s blocking the door. He says, “No, I have 
instructions to stay here.” She says “From who?” He says “From the doctor.” 
The nurse goes off and gets the head nurse, who asks him “Why do you 
refuse to move from that spot?” He says “Because the doctor told me to wait 
here.” The head nurse, as he notes, “couldn’t assume the risk of overruling 
the doctor” so she leaves and returns a few minutes later with her supervisor. 
Here is the conversation with the supervisor. 

“Which doctor,” she asked, “told you to stand there?” 

“I don’t know, he didn’t tell me his name.” 

“How long ago,” asked the supervisor, “did the doctor tell you to 
stand there?” 

“It must be about five hours ago,” I replied. 

“Do you expect us to believe that a doctor whose name you don’t 
even know told you to wait there five hours ago and that he hasn’t come 
back yet?” 

I remained silent. The question was extremely sharp, and I winced 
under its impact. 

“Are you sure that the doctor is not just a figment of your 


Part I 

I must admit that my confidence was shaken. I began to doubt, 
myself, that this had happened. Plainly doctors don’t break their 
promises. If one of them tells you that he’ll return in five minutes, he 
won’t keep you waiting for five hours. Such things just don’t happen. 

“I must be crazy,” I said weakly, “Maybe it didn’t happen.” 

The supervisor nodded. “That’s better,” she said grimly. 

My head reeled and my legs tottered. I was beginning to feel like the 
prisoner who is kept in a padded cell for years without ever being 
informed of the charges against him in Kafka’s novel The Trial. 
Everything was turning topsy-turvy. Guilt was innocence and innocence 
was guilt. Nothing admitted of any rational explanation. If that doctor 
who had told me to wait for him for five minutes were merely a figment 
of my imagination, then I was losing all contact with reality. But if that 
doctor really had existed, then why did all these people maintain that he 
didn’t? I knew that he existed. I had seen him and spoken to him. Then 
all these people must be engaged in a monstrous conspiracy to drive me 
mad. No other theory seemed to fit these facts. 

The supervisor interrupted my thoughts. “Come with me,” she said. 

I obeyed meekly and followed her towards the center of the hallway. 

“You're very confused, young man,” said the supervisor. “Because 
you’re new here I won’t report how you’ve just behaved. But I warn 
you, don’t do it again. I want no further trouble from you.” 

Again, then, we see the use of this procedure - that one can choose among 
facts according to the presence or absence of an explanation. It’s absolutely 
routinely used. I’m not proposing that it’s obscene, I’m only proposing that 
that’s the way it’s done. At least in this society, facts and explanations have 
more than a one-way relation to each other. That is, it isn’t the case that if 
something has occurred, that sets the problem ‘construct an explanation,’ but 
the notion that persons hold of possible facts is that those facts are possible for 
which there is an explanation. One can’t merely say “Well, I saw it. You 
explain it.” Something proposed to have occurred can be treated as not so, by 
virtue of the fact that there’s not an explanation for it. That’s important in this 
society, given the fact that miracles are no longer usable. And miracles are that 
class, in part. They’re events for which there is no account, which now 
systematically would be given an account, i.e., an account not of this world. 
But most persons who consider themselves to be modern individuals don’t 
read, for example, the psychical researcher, Rhine. They just take it that 
whatever he says, he’s a fraud in one way or another, and you don’t have to 
bother coming to terms with what he reports. Given that what he proposes 
as an explanation couldn’t be an explanation, then it hasn’t occurred. And 
further, there is an explanation available which provides that there is no such 
phenomenon: Statistical chance. Such an explanation proposes that out of a 
population there will be some persons who produce these responses, and he’s 
happened to have found them. But that does not provide any basis for psychic 
research; random distributions would do the same job. 

Lecture 14 


That’s a classical problem. It can always be raised if you have a limited 
amount of research, where, then, if you propose there’s a phenomenon, it is 
said that you have nothing but the working of a random distribution, which 
you happen to have caught at some point. And it’s raised more than 
occasionally about psychic research, and I suppose more than occasionally it 
turns out to be relevant. Or, for example, people say “How does it happen 
that there have been two earthquakes on this day?’’ and then they go about 
constructing some explanation: The gods are angry. Or, “Isn’t it odd that the 
weather pattern has changed this year?” and then they construct an 
explanation: Fallout. And now what’s proposed is that if you use such a unit 
as ‘years’ to measure weather, where weather as a phenomenon has to be 
measured over eons, then you’re liable to find something which looks like 
order. But given the proper unit, that proposed order is not present. Now, 
there are all sorts of units available, and for any proposed phenomenon, that 
unit which would make that proposed ordered set of facts simply a 
coagulation of random events can be proposed to be the relevant unit. Either 
side may be right or wrong. Nonetheless it’s done. 

Appendix A 
A Note on the Editing 

In lecture 2, p. 17, Sacks introduces a consideration: 

A recurrent thing that I’ve seen throughout this stuff is persons talking 
about not feeling wanted anymore. The question is, how is that kind of 
feeling provided for in this society? And what would be interesting 
about it would be if we could see some way in which, quote, the 
structure of society, provided for the focusing of kinds of troubles. 
That’s what I think we can see with this . . . 

Throughout this volume many of Sacks’ pronominal uses have been changed. 
Here, the operation is more or less innocuous. What is rendered as ‘. . . if we 
could see . . .” and “That’s what I think we can see . . .’’ actually goes, 
“. . . if we could see . . .” and “That’s what I think you can see . . . i.e., the 

second ‘we’ is actually ‘you.’ Following is the unedited version. (Emphasis on 
pronouns is always added: it is not in the transcripts.) 

Now, a recurrent thing that you have probably seen, and I’ve seen 
throughout this stuff, is persons talking about not feeling wanted 
anymore. The question is, how is that kind of feeling provided for in 
this society. And what would be interesting about it is, if we could see 
some way in which, quote, the structure of society provided for the 
focussing of kinds of troubles. That’s what I think you can see with this. 

(Fall 1964, tape 5, side 1, p. 10, unedited) 

And this ‘we’/‘you’ alternation is preserved in the abbreviated earlier 
transcript (transcriber unknown). 

The recurrent thing is people talking about not feeling wanted. How is 
that kind of feeling is {sic] provided for in this society? If we can see 
some way in which the structure of society provided for the focusing of 
certain kinds of troubles. That’s what you can see above. 

(Fall 1964, tape 5, side 1, pp. 3-4, transcriber unknown) 

Another ‘you’/‘we’ alternation occurs just a bit further in the discussion. Here 
is the unedited version. 

Now you see that what she’s picking up here is, quote, the absence of 
ceremonials. Now, we see that ceremonials have this double use. 

(Fall 1964, tape 5, side 1, pp. 10-11, unedited) 



Appendix A 

(Neither the edited version nor the earlier transcript show the repeated, 
pronoun-alternated reference to ‘seeing.’ The edited version has it: 

What she’s picking up here is the absence of ceremonials. And 
ceremonials have this double use. 

(Fall 1964-Spring 1965, lecture 2, p. 18) 

And the earlier transcript goes: 

She picks up the absence of ceremonials. Ceremonials have a double use. 

(Fall 1964, tape 5, side 1, p. 4, transcriber unknown) 

So, in close order there are two of these pronominal alternations: “. . . if we 
could see . . . ‘/’I think you can see” and ‘‘Now you see . . .’’/“Now, we 
see ...” And such alternation is recurrent across the lectures. Following are a 
few instances, taken from the Fall 1964 and Spring 1966 unedited lecture 

Now, what you can see is this . . . And what we find is . . . 

(Fall 1964, tape 1(R), p. 6, unedited) 

This cross-paragraph alternation survives in the edited version (Fall 1964- 
Spring 1965, lecture 1, pp. 6-7). 

If you look at [X and Y} what we seem to find in them is . . . 

(Fall 1964, tape 4, side 2(?), p. 1, unedited) 

This alternation is edited out and rendered as: “What we find in [X and Y] 
is . . .” (Fall 1964-Spring 1965, lecture 7, p. 54). 

I could say [X is the case] but / wouldn’t be able to show why that’s so, 
or how it’s so. If that’s the case, I don’t establish my point. I could say 
[Y] and we could [find A but not B}. That again wouldn’t do what we 
propose to do. We want to do both: [A and B], You want a method that 
generates this. 

(Spring 1966, lecture 4, pp. 1-2, unedited) 

In the edited version, a ‘we’ referring to ‘I’ is changed to ‘I’ (“That, again, 
wouldn’t do what I propose to do”), and a ‘you’ following a prior ‘we’ is 
changed to ‘we’ (“We want to do both: [A and B ] ... we want a method that 
generates this line. (Spring 1966, lecture 4, p. 301). 

In short, what you want to do is [X]. And we can take it that [Y], 

(Spring 1966, lecture 21, p. 6, unedited) 

This alternation survives in the edited version (Spring 19 66, lecture 21, 
p. 420). 


Part I 

If you can claim that it could be another thing, first one wants to show 
how it’s another possibility. 

(Spring 1966, lecture 22, p. 2, unedited) 

This alternation is revised to, “If one is claiming that it could be another thing, 
first one wants to show how it’s another possibility’’ (Spring 1966, lecture 21, 
p. 422). 

Many of the pronominal alternations are edited out or revised. One major 
basis of revision is an attempt to deal with the ‘direct address’ problem. 

Sacks often addresses remarks directly to the class, as occurs in the unedited 
transcript of the paragraph in question here: 

Now, a recurrent thing that you have probably seen, and I’ve seen 
throughout this stuff . . . 

He also makes such meta remarks as, “I would like very if you could collect 
instances of the uses of {X}” (Fall 1964, tape 4, side 2, p. 11, unedited). 
‘. . . if it isn’t clear, stick up your hand and tell me’’ (Fall 1964, tape 1, p. 2, 
unedited), not to mention things like “Oh by the way. Somebody tell me 
when I get - I don’t have a watch. So just keep me vaguely informed about 
( )” (Fall 1964, tape 1, p. 9, unedited). These have been edited out. 

But also, lecture-relevant talk which was addressed to the class was often 
deleted or changed because it could be troublesome to readers, either 
estranging them, making them onlookers or overhearers, or over-intimatizing 
the talk, seeming to address this particular reader directly. So, e.g., such a 
remark as, “You ought by this time to be quite aware of the fact that 
(Spring 1966, lecture 6, p. 1, unedited) was changed to “By this time it 
ought to be quite obvious that . . . ’’(Spring 1966, lecture 6, p. 312). In the 
actual situation, of course, members of the class were neither overhearing, nor 
being individually addressed by, such utterances. 

There are, then, variously problematic uses of ‘you.’ And the fact that 
Sacks not infrequently uses ‘you,’ ‘one,’ and ‘we’ in alternation became a 
resource - and license - for changing them. Most often ‘you’ was changed to 
‘we,’ a word Sacks makes extensive use of, sometimes in rather idiosyncratic 
ways. So, for example, he will use ‘we’ where he himself is the referent, e.g.: 

That again, wouldn’t do what we propose to do. 

(Spring 1966, lecture 4, p. 2, unedited) 

Let’s start out with something we mentioned before . . . 

(Fall 1964, tape 14, side 2, p. 2, unedited) 

Remember we said about the opening lines of conversations that they 
seem to come in pairs. 

(Fall 1964, tape 5, side 1, p. 3, unedited) 


Appendix A 

Now we originally introduced the notion of an omni-relevant device . . . 

(Fall 1966, lecture 4, p. 6, unedited) 

in each case it being he himself who has “propose{d},’’ “mentioned,” “said,” 
and “introduced [a] notion.” 

In one case, Sacks recasts his own utterance: 

We haven’t said that they have a right to be told. I have not said that 
they have a right to be told some trouble of A’s. 

(Fall 1964, tape 10, side 2 (M3 side 2), p. 11, unedited) 

And in the edited versions, either ‘we’ is changed to ‘I:’ 

That, again, wouldn’t do what I propose to do. 

(Spring 1966, lecture 4, p. 301) 

I’ll start out with something I’d already worked up . . . 

(Fall 1964-Spring 1965, lecture 8, p. 57) 

I said about the opening lines of conversations . . . 

(Fall 1964-Spring 1965, lecture 2, p. 13) 

or the talk is revised to accommodate ‘we:’ 

We have now, the notion of an omni-relevant device . . . 

(Spring 1966, lecture 6, p. 316) 

Sacks also uses ‘we’ to refer to pretty much anyone but himself, e.g.: 

Hopefully the thing I want us to see in this first is . . . 

(Fall 1964, tape 5, side 2, p. 1, unedited) 

I raise this because while we all can see that that’s quite so, there’s a 
related and in a way much more interesting thing that I doubt we’ve 
noticed . . . 

(Spring 1970, lecture 1, p. 7, unedited) 

We tend somewhat to be perhaps overly taken with the constraints that 
Weber sets for an objective social science . . . 

(Spring 1966, lecture 33, p. 9, unedited) 

where it is members of the class, and not Sacks, who will “hopefully” come 
to “see” something, and whom he can “doubt” have “noticed” something. 
And it may be an even larger population, but in any case one that doesn’t 
include Sacks, whose members “tend to be perhaps overly taken with” 
Weber’s constraints. 


Part I 

While these self-inclusive, ‘we’-references may be incorrect in a strict sense, 
they may be doing the sort of work Sacks discusses in one of the lectures; that 
of saying ‘everyone.’ 

If someone who needn’t include themself in some class . . . wants to say 
in a stronger way than ‘you’ (which can be misheard as only the 
recipient) that ‘everyone is that way,’ they can say ‘we.’ 

He offers a case in which a psychiatrist remarks, “We can be very blind to the 
things around us,” and goes on to say: 

That is, by virtue of the sequence in which ‘we’ is heard, the inclusion 
of the speaker is partially gratuitous; is heard as doing something he 
‘needn’t have done.’ 

(Spring 1966, lecture 11, pp. 351-2) 

As it happens, in the edited versions those instances of “gratuitous” 
self-inclusion in the environment of a demand, “. . . I want us to see . . .,” or 
criticism, “. . . there’s ... a much more interesting thing that I doubt we’ve 
noticed” and “We tend somewhat to be perhaps overly taken with [X]. 
were preserved. (See Fall 1964-Spring 1965, lecture 3, p. 21, Spring 1970, 
lecture 1, p. 00, and Spring 19 66, lecture 33, p. 487.) 

In contrast to those pronominal changes mentioned earlier, both the 
changes from ‘we’ to ‘I’ when Sacks himself is the referent, and the 
preservation of the self-inclusive ‘we’ in the environment of touchy assertions, 
were not matters of policy; they were decided upon one at a time. But they 
can be seen to add up to something like policies, or perhaps to reflect 
underlying policies. 

The paragraph at the start of this consideration, plus a bit of subsequent talk, 
happen to instance another phenomenon that was often revised. The unedited 
transcript shows a format, “quote, [X},” which occurs once, 

... if we could see some way in which, quote, the structure of society 
provided for the focussing of kinds of troubles . . . 

and is then used again some 16 lines later. 

. . . what she’s picking up here is, quote, the absence of ceremonials. 

(These two fragments are from Fall 1964, tape 5, side 1, p. 10, 11. 4-6 and 
p. 10, 1. 23-p. 11, 1. 1, respectively. See p. 126 of this Appendix for fuller 
fragments of each.) 

This phenomenon of a repeated format occurs throughout the lectures, and 
often - as here - one of them, usually the second, is deleted in the edited 

Appendix A 131 

version, for both aesthetic and technical reasons. Only the latter will be 
considered here. 

One sort of technical problem posed by the repeated format is that a term 
which may appear to be analytically based, can be merely format generated. 
And a glimpse of that problem may be gotten from the recurrence of the term 
‘quote’ in “. . . quote, the structure of society ...” and ”... quote, the 
absence of ceremonials.” Roughly, the quote-marking of the familiar phrase 
‘the structure of society’ seems to be doing its standard ‘as they say’ work, 
where a new way of working with an old concept is being offered. It is not 
clear what sort of work the quote-marking of the newly minted technical 
descriptor ‘the absence of ceremonials’ is doing. One could not in this case 
substitute ‘as they say’ for ‘quote.’ It may simply be a format-recurrence. In 
any event it was treated as such, and deleted from the edited version which 
simply goes: “What she’s picking up here is the absence of ceremonials” (Fall 
1964-Spring 1965, lecture 2, p. 18). 

Part II 
Fall 1965 

Lectures on Conversation, Volume I, II Harvey Sacks 
© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6 

Lectures 1 and 2 

“The baby cried. The mommy picked 

it up.” 

Lectures 1 and 2 were not recorded , but Sacks had made preparatory notes for the 
“first lecture(s)" which seem to have turned out to cover lecture 1 (see Appendix 
A). The second lecture is covered in notes Sacks made after its delivery (see 
Appendix B). 

Lectures on Conversation, Volume I, II Harvey Sacks 135 

© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6 

Handout Group Therapy Session 


The door opens in the middle of a group therapy session, and in comes a new 
entrant. Until now, at this session, the members have been three teen age 
males, and a male therapist of around 35 years of age. The new entrant is a 
male teenager. 


0:00 Therapist'. 












r Mel 

- Henry 
L Henry 


Bob this is uh Mel 








Bob Reed 

(cough) We were in an automobile discussion, 
discussing the psychological motives for 
drag racing on the streets. 

I still say though that if you take, if you take uh a big 

fancy car out on the road and you’re hot roddin’ around 

you’re you’re bound to get, you’re bound to get caught 

and you’re bound to get shafted. We- look 

Now did you do it right. That’s the challenge That’s the 

challenge You wanna try 

That’s the problem with society. Hahhh 

And do it right so you do not get caught. 

That’s the 

In that Bonneville of mine? I could take it out with me 
and if I got a tie and a sweater on and I look clean? 99% 
of the time a guy could pull up to me in the same car the 
same color, the same year, the whole bit, roll up his pipes 
and he’s in a dirty grubby tee shirt and the guy will pick 
the guy up in the dirty grubby tee shirt before I before 
he’ll pick me up. 

Just - just for - 

This is as close a copy as possible of Sacks’ transcript. 














1:30 Henry 



2:00 Bob 

Handout: Group Therapy Session Segment 137 

Not many people get picked up in a Pontiac station 

Now I agree it looks like a daddy’s - It looks like a damn 
mommy’s car. 

Joe, face it. You’re a poor little rich kid. 

Yes Mommy. 

Face the music 

Ok Now you’ve got that out of your system. Now you’re 
a poor little rich kid we’ve told you that. 

And we also decided you’re a chicken shit. 

I decided that years ago. Hell with you 

Now let’s see what else can we decide about you? 

Hey don’t tear him down 
I’ve been torn down for- 

We got company 

oh ok. Tell us all about yourself so we can find something 
bad about you. 

Yeah. Hurry up. 

Well first of all you must be crazy or you wouldn’t be 


heh heh 

Yeh I guess 

Secondly, you must be an under-achiever 

You hate your mom and dad, huh 
Third of all — 
oh sometimes 

Fourth, you like to drive cars fast, 

Fifth you like uh you like wild times 

He smokes like me see. 

Sixth, you like booze 

Seventh you like to smoke 

And seven you’ve been arrested for rape and other things. 


Eight you- 
No. not that. 

Eight you Eight you give lip back to everybody. 


Nine you uh cut classes, 

Ten, you’ve been kicked out of school once, 


Part II 












2:30 Mel 

3:00 Henry 











eleven uh 
hah hahh 

We’re doing better than he is. haha. Proceed, I’m with 
you. ha ha. 

Your turn. 

Are you just agreeing because you feel you want to or 



You just agreeing? 

What the hells that? 




Agreeing. With us. Just going along with it. 


Saying Yes. Yes. hahhh 

Its true, heh, everything you said is true. so. 

You wouldn’t be in this group if you didn’t. 

What school did you go to? 

I went to Palisades 

Yeah. And then University. 

Oh I went to Uni. I graduated there. 

You did? 

Bitchin school isn’t it 
Yeah I guess 

Yeah I never did ( ) it 

What did you do with the matches? 

You don’t like it do you? 

I got them 
Here you go chum 
Everybody’s faster than me today 
You got reflexes like a slow turtle 
God damn you ha 

Remind me later. I’ll show you a litte game about your 

Got any more cigars. 

No it’s last one 
Last one 
Pity heh 

Usually there’s a broad in here. Pretty - 

Her name’s Barbara. She pretty good. She’s real nice. 


A lot of her to love but 
Sounds like an old man 

Huh. This is the old, this is the couch society of what is it? 




3:30 Mel 


















4:00 Henry 












Handout: Group Therapy Session Segment 

Wanna join our army? 

Wanna join our army? We’re gonna go defeat the king of 

We’re gonna 

The academic counselling center fightin’ army’s gonna go 

Alright Mel, he’s been here ten minutes 

Mel. Your match is still going Mel 

Mel He’s been in here ten minutes let’s judge his 


ok. ready Henry? 


You’re first 

You’re gonna get killed 

Keep your mouth shut and don’t say a damn word, ha ha 
He’s not at all like Joe. 


He’s more like Mel and I 

( ) 

What you refer to as - hippie, ha. He’s been in it up to his 



A couple of times 
Ha ha. Taste good? 

No. I bit off the end of it. I was chewing the end of it. 


Havin a big hassle with your folks 

Right. Daddy wants to keep you down Mommy 

Mommy you can tell what to do 
(cough) God damn cancer 

See. I told you, we know everything about this guy already 

Do you have any brothers or sisters? 


Daddy’s sort of sheltered you 

No not really 


Not really 

You sorta shelter them? 

Do you have a car? 



Part II 


4:30 Mel 




| Henry 

r Henry 






5:30 Bob 





L Mel 

oh oh. One point against you. 

You had a car? 


You want a car 
oh hell yes, ha. 

You’ve stolen a car? 

Daddy doesn’t want you to have one. 

No he wants me to have one, my old lady doesn’t. 

oh!! Main conflict with mother. 



Mother’s too dominating, 

But you’ve been had already fella, ha ha 

Bring your mother in, we’ll work on her. ha ha 

What are you thinking, old man? 

hahh I’m thinking Bob has certainly had it. hahh 

Whip out your pad 

Five minutes and you guys 

No No write down what we say take the ( ) 

You gotta pardon these guys they ’ re all they ’ re all the 

social outcasts of the world. 

This one’s chicken shit. He’s a bastard, and I’m a Mau 

ok. good. Now we’re all settled, what are you? 
oh I’m 
You a hood? 

This is an abnormal session. See we’re not together without 

the broad. 


See we gotta have the broad here. Cause she unites us. 

She unites us - she keeps us on the road. Ha ha. Table’s 
always ready and ha — 

She sorta keeps us on the road 
oh well 

oh well you know that the way things go. That’s the way 
things lie down. 

He sorta keeps his mouth shut and writes down things 
whenever you say something important 
And then if you ask him 
He’s a good guy though 

to do something then you have to pay him though 





6:00 Henry 


Handout : Group Therapy Session Segment 

Eventually you’ll become sane, 
yeah but you — 

Or your mother will which is not what you think. 

I hope so 

You’re not crazy, it’s your mother, hah? 

But uh it takes about, I’ve been here four months and it’s 
getting better now. I used to think that I was a bird and 
would fly. 

Yea He used to walk out on us, he thought he was above 








6:30 Henry 







7:00 Joe 








Yea. But now I’m now I’m below you. 

Yeah. I corrected I corrected that quality. I gave him an 
inferiority complex, ha. 

And I got him to shave, 

Yeah. I’m not grubby or nothin 
No. hehh 

hah. Hey this is the academic counselling center. It’s called 
the family, family circle. 

It’s not really an academic counselling center; it’s sort of a 
drive in nut house, ha ha. 

Then you rent a counsellor; $ 5 an hour and 50<t a question, 
ha ha. 

Don’t consider this one- but on the whole you’ll enjoy 
’em. Ha. In fact it’s a good way to spend Saturday morn- 

After Friday night. 

The morning after the night before. Ha ha. After a couple 

cups of coffee you’re alright. 


A couple dozen cups of coffee. 

Chew a cigar now and then and 
I don’t smoke (he’s been smoking a cigar) 

He just started last night 

You gotta believe it. You gotta believe it we’re the only 
two in here that smoke other than him. 

Doesn’t Barb smoke? 

No. she doesn’t smoke. She said she used to. 

She’s a good girl, ha ha 

She smokes cigars. That’s about all. ha ha. 

They smoke. I chew them. 

He eats it. 

Well chewing them is half the fun - oral gratification, heh 


Part II 




7:30 Henry 












8:00 Henry 











8:30 Joe 

Let’s see, what else can we tell him? 

They say that’s why bubble gum and all that shit’s good uh 

so popular 


There’s a trend to put things in your mouth - like thumbs 

No really. Why do you think babies suck their thumb and 
adults chew gum and smoke? Everybody wants to keep 
something in their mouth. 

Well, you know - 

Now I’m not talkin about eating or nothing. And I’m not 
talking about what you’re thinking of either. Hahh. 

No really I read this in some book 


oh come on 

It’s really the truth 

one more thing — 

Look at him. Look at him sucking at that pipe all goddam 


He sucks on his dentures too. Now listen, one more thing. 
There’s people watching us now. They got a tape recorder 

Thai is the microphone (pointing). Occasionally 
there’s somebody in there judging us. They’re baby 
headshrinkers. They’re learning how to be. 

Don’t worry about ’em. They destroy the tapes afterward. 
He’s not quite one. He won’t be one until two months. 
He’s writing a - 

paper. But today we’re not being uh viewed. We’re just 
being taped. 

No. we’re being viewed. There’s somebody in there. Cause 
they lit a cigarette. 

But he’s blind. 

Hey. Don’t worry about it. 

How how could you see him? 

Well he can see us but you can’t see 
Can you see through this here? 

When you turn the lights on. See it’s darker in there. I 
think - 

Wave at him. He’s a nice guy. 

Anything else Henry? 

And if it gets private we can always shut the curtain, 
heh great, 
to do something. 













9:00 Henry 

Handout: Group Therapy Session Segment 143 

Anything else Henry? D’you think we oughta let him in 
on? or tell him? 

We’re all nuts? 

Naw. Let him slide for a few weeks, 

Get the hang of things, heh. 

Then we’ll really go after him. Then when we - when we 

get - 

We ll let you rest with your problem now 

We’ve told you all your problems now you gotta rest with 



Then we’ll tell you how to solve them. 

You gotta get warmed up 

And vice versa. 1 

1 On the original transcript, the materials fitted precisely onto six pages. The tape 
goes on for another hour and 40 minutes, or so. Whether the fragment would have 
continued on until the bottom of a next page, or have cut off at this neat 
time-and-topical place, is an open question. 

Lecture 3 

A collaboratively built sentence; The 

use of ‘We’ 

I’m going to begin discussing this eight-minute segment from a group 
therapy session for teenage kids that I gave out last time. When I used this 
last year as a final exam, I offered as a suggestion for considering it, that I took 
it to be an initiation ceremony, and as such, among the things in it that are 
of interest is the fact that it isn’t done via a script. I think it may be the case 
that one of the participants in the group knew beforehand that somebody new 
was going to come that day, but that’s about it. There isn’t any basis at all for 
supposing that this thing was planned by them. And yet (if ‘yet’ is at all a 
relevant term; it’s relevant vis-a-vis the literature on initiation ceremonies 
anyway) there are certainly features in it which are proposed to be present in 
initiation ceremonies. But our task is not going to be so much to see the extent 
to which that is so, but to try to come to terms with this thing, as best we can. 

I’m going to begin consideration of it with what I take to be Segment 2, 
which runs from the cough which is right after the introduction of the various 
personnel, until that point which is very close to the 1:30 marker, where Mel 
says “Oh, ok.’’ So it runs from “cough” to “oh, ok,” inclusive. That’s going 
to be the subject-matter for these first considerations. Perhaps we’ll eventually 
move backwards and forwards. 

There are, in the conversation itself, a lot of events that are to the altogether 
naive eye, quite remarkable. That is, without any analysis and simply by 
inspection you can find some things that you might take to be worth thinking 
about, without any special consideration of what we’ve done here at all. 
Probably the most striking thing right off, would be that part of the third 
section where there’s a series of eleven questions run off in about 45 seconds, 
by the given members of the group to the new entrant, in which they ask him 
things about himself, giving in each of the questions what they take to be the 
answer. And where the questions have some rather extraordinary detail in 
them, for example, question ten, “Ten, you’ve been kicked out of school 
once.” The issues of how these guys have that information, and what they do 
with it, are quite obvious ones. And perhaps we’ll eventually get to consider 

Segment 2 begins with what is, in its way, also a rather remarkable 
occurrence; and that is, the three boys collaborate to produce a single sentence: 

Joe : (cough) We were in an automobile discussion, 

Henry : discussing the psychological motives for 


Lecture 3 


Mel : drag racing on the streets. 

It’s not altogether unheard of for two persons to collaborate to produce a 
single sentence. The normal way that is done, however, is that, say, one person 
produces an almost complete sentence and finds himself searching for a last 
word or a last phrase which he can’t find, and the other offers it. Another place 
where one gets something like this is in the following sort of thing: A person 
makes a request - ‘Would you pass the salad.” Somebody else, maybe the 
person who got the request, then appends to the request some word or phrase 
that they propose properly completes it. . .((voice from the gallery: “Please”)) 

. . .“Please;” which they propose that, lacking that, the request is not 

Now those are two sorts of things worth considering in their own right. But 
they’re far off from this thing we’re looking at. It’s a kind of procedure of such 
power that if it were used as a literary technique it would just be rather vulgar, 

I suppose. I know a couple of literary uses, though I can’t really pin them 
down for sure, and if anyone knows of any you could tell me sometime, or 
collect them yourself. 1 In a musical version of Edgar Lee Master’s Spoon River 
Anthology he uses this as a way that the pair of lovers talk. 2 That is, the pair 
of lovers talk in the following sort of way: Each one produces part of a 
sentence, which the other then may complete. So you write out a sentence, 
break it somewhere, and have the other part of the pair continue it. 

Now Masters was writing a while ago, and I take it that we would take it 
that that’s an altogether obvious device to show, through this playing with the 
syntactic features of an utterance, that these people are close to each other. 
They’re a unit. Because a sentence is obviously a prototypical instance of that 
thing which is done by some unit. Normally, some single person. That then 
permits it - for those who have the wit to do it - to be a way that some 
non-apparent unit may be demonstrated to exist. 

We get, then, a kind of extraordinary tie between syntactic possibilities and 
phenomena like social organization. That is, an extremely strong way that 
these kids go about demonstrating that, for one, there is a group here, is their 
getting together to put this sentence together, collaboratively. 

It’s hard to figure how they could do that right off, in anything like as 
sharp a way as they picked. As it happens, as rare as I take it that this kind 
of a thing is - and until I saw it I don’t think I had seen anything like it before 
- in the very same segment there’s another. Between 1:00 and 1:30 we get 
the following sequence: 

Mel : Ok Now you’ve got that out of your system. Now you’re a poor 

little rich kid we’ve told you that. 

Henry : And we also decided you’re a chicken shit. 

1 Written in at the margin: “Donald Duck. Huie, Louie & Dewey.” A student 
contributed this. 

2 Written in at the margin: “cf. Freud.” 


Part II 

In the case of the participle “discussing,” which constructs what is being 
said as a dependent clause and makes, then, the statement before it a clause 
also, the phenomenon is rather sharper than this one, where “and” is a rather 
normal way to begin a sentence. But it seems to do exactly the same kind of 
job, for, now a different group. That is, if the first group (Joe, Henry, Mel) 
is formed through their production of “We were in an automobile discus- 
sion . . . ,” etc., and is demonstrated to be present through that production, 
as an organized entity in this terribly powerful sense of being organized in 
such a sufficient way that a task properly done by a single person is done by 
all of them together - and not as a chorus - then, there are clear ways that 
Mel’s “Now your’re a poor little rich kid we’ve told you that,” followed by 
Henry’s “And we also decided you’re a chicken shit” is very similar. I’ll 
consider this one when we get down that far into the conversation. And if you 
examine the conversation with the slightest attention, certainly one would 
take it that at that point in it a subgroup is being formed, consisting of the 
two people (Mel, Henry) who collaborate on that sentence. 

Such mobilizations of syntax are normally assigned to purely literary 
phenomena - poetry, the novel, the play - as their expectable environments. 
That they might be explored by laymen doing quite unformalized tasks is 
something that is, I take it, quite worthy of being investigated - for other 
things than just this particular way that these syntactic features are being 

Now let’s see to some extent how that thing is done. Joe’s utterance, “We 
were in an automobile discussion,” is something which by itself could be a 
perfectly decent sentence. And it is, then, in the first instance, what Henry 
does that makes it the independent clause of a sentence which is now to have 
a dependent clause. That is, Henry makes his own statement not possibly a 
sentence by using the participle “discussing,” which is just not a way of 
beginning a sentence . 3 

About the third part (Mel’s “drag racing on the streets”) there is no 
question that it collaborates with the second in making of the first, ‘the 
independent clause.’ Neither the second or third alone are sentences, and the 
two together do not make a sentence. Only with the first is it all a sentence. 
So that particular choice of participle is to be accounted for by reference to 
some task of social organization, solved by reference to syntactical features. 
And not by any, quote, purely linguistic considerations. Or even stylistic 
considerations. The participle, then, becomes an object in the technology of 
social structures, I suppose. And we get some work that it can do which one 
isn’t going to find much in whatever radical grammar you’re going to look 
at . 4 That would be trivially true for almost every grammar, since they would 
never make a distinction in the first instance, between speakers. And yet it’s 
obvious that, that this statement is made by three persons makes it quite a 
different statement than if it was made by one. And if you took it as all made 

3 Written into the margin: “Needs to be modified, for it can begin a sentence.” 

4 Written into the margin: “And we don’t find this in sociology texts either.” 

Lecture 3 


by one, the various ways that they might have of doing the kinds of things 
they can do when they make it together, would be enormously weakened. 
Suppose, for example, Joe had said the whole thing. And the two of them 
either said “Yeah,” or nothing. The character of what they would be 
affirming would be possibly obscure. 

So it’s not only that they present themselves and demonstrate that they are 
organized, but by collaborating on the sentence they have this extremely 
powerful way of showing that they do agree as to what it is that the topic was: 
They all have some part in saying what it was. And I take it that that’s a 
different kind of technique of affirmation than is, A says the topic was X, and 
B says yes it was. And that would be especially relevant where the question 
is, what does that topic mean; where, for example, a person who hears it may 
take it that he is not sure, and perhaps neither of them really agree on it. 

In the way that these fellows fit their talk together, which would involve 
a hearer in seeing that they ‘know what’s on each other’s minds,’ he could 
presumably as well take it that what it is they’re talking about, they also 
know. And that the thing works this way would of course be further 
evidenced if we had more studied uses of it; that is to say, literary uses, for 
example, where it is obviously just that point that’s being made. That is to 
say, what better way is there for lovers to show that they are one . 5 

In that regard it’s perhaps worth opening up one’s notion of what it is that 
organization might be directed to doing. There is, I take it, a considerable 
tendency to think of the task of organization as being the solution, by a set of 
persons, of some job which no single one could, or could efficiently, do. 
Although that is certainly so to some considerable extent, the fact that there 
is a job that any person could clearly do by themself, provides a resource for 
members for permitting them to show each other that whatever it is they’re 
doing together, they’re just doing together to do together. That is to say, if one 
wants to find a way of showing somebody that what you want is to be with 
them, then the best way to do it is to find some way of dividing a task which 
is not easily dividable, and which clearly can be done by either one alone. And 
that, I take it, obviously is done rather frequently, if not with the putting 
together of sentences, then with all sorts of other things. Where neither party 
can then readily take it that the other is simply being of help, but is involved 
in seeing that what they’re doing is ‘doing whatever you’re doing with you.’ 

Now there’s certainly ambiguity in ‘divisible’ as a term. And what I’m 
pointing to is not something which is not readily divisible and is also 
unmanageable by one person, but something which is not readily divisible but 
which is manageable by any single person. The point can be made by 
paraphrasing a fantastic aphorism of Kafka’s. He loved to play with the 
problems of social organization, and give them queer formulations. And at 
one time he’s considering Neptune’s problems. Neptune is the bureaucrat 
who runs the oceans, according to Kafka. And he has the following kind of 
problem: How in the world is he supposed to divide up that kind of work? 

5 Written into the margin: “cf. Freud” 


Part II 

You can’t, after all, assign somebody an ocean. It’s not that sort of 
non-divisibility that we’re talking about. 

Now there are, aside from what I’ve just mentioned, a variety of other 
things involved in this segment that we’re going to have to consider in one 
way or another. 

An extremely sharp formulation is made of the term ‘we,’ where it remains 
the subject of the sentence without each person affiliating himself to it. And 
we might then say about ‘we,’ that what it refers to is ‘those three guys.’ That, 
however, would be a possibly troublesome kind of way of characterizing it. 
For what we might have to decide, then, is what kind of form of ‘those three 
guys’ does it refer to? That is, is it a list in which they happen to be the only 
people on the list - that is, directly pointed to by ‘we? They talk in a sequence 
which might permit one to say they are a list of persons, and previously they 
are introduced as a list; a list of nameables. Now if it is a list, then there are 
of course a variety of kinds of operations which might be permissable on that. 
For example, order might not matter at all. If we wanted to write them down 
we could say ‘we:’ (Mel, Henry, Joe) or (Joe, Mel, Henry), it doesn’t matter. 
As long as you have them there, that’s it. At least for some purposes it would 
seem intuitively obvious that order is not irrelevant to the use of those names, 
and if we go back to the first segment, we can see such a one. 

When the new entrant comes in, the others are sitting around the room, in 
chairs. There is the therapist, then Mel, then Joe, and Henry. Bob comes in, 
and he’s introduced. There’s a pause in the introduction, right at the 
beginning: “This is uh Mel.” Now we can take it that at that point matters 
are open as to how they’re going to be introduced, i.e., in what sequence. And 
we can take it that probably only two alternatives are present: Go around to 
the therapist’s left, or around to his right. But you wouldn’t start in the 
middle, i.e., with Joe. Once the therapist has named Mel, presumably 
anybody who came in and knew that Mel had been introduced, could do the 
job of introducing. And if it were done in any other sequence, we would take 
it that something special is being done. For purposes of introduction, then, if 
the persons are in some way physically arranged with respect to each other, 
then we would take it that it isn’t simply an open matter as to what sequence 
they ought to be introduced in. There may be some openness to it, but 
probably, given the first name, there’s a proper further sequence to the end. 

Now the problem about ‘we’ that we have to deal with, which is only 
roughly posed by the issue of orderability, is the following. Is it the case that 
‘we’ is some collection of these guys’ names, directly? Or is it some category 
or a set of categories for which these fellows are incumbents? Where, then, 
their names are usable to refer to them, but there is some organization, 
perhaps, of those in that category. Where that organization is what’s 
important. And saying that, then we have to turn to it first. Do we have to 
build a category here? Or can we turn simply to the list of names as the thing 
being referred to? 

‘We’ clearly can refer to a category, which has as one of its crucial properties 
that no intention exists of listing the incumbents, and furthermore they’re not 

Lecture 3 


listable. That is, ‘we’ can refer to an infinite population. For example, 
‘Americans.’ That’s an obvious infinite population. People on a bus are an 
infinite population, insofar as you don’t say ‘at time T.’ ‘We’ is stable over the 
use of ‘Americans,’ and therefore it might be only incidental that these people 
are listed. 

Now, “What’s the difference?’’ may be a question for you. If it were the 
case that ‘we,’ in this instance, is properly used as the substitute for a listing 
of the personnel and could not be used unless a list could be made, and a list 
were intended, then it’s clear that ‘we’ must get a tremendously more 
restricted use than it in fact has. 

One way we could go about approaching the issue of, is there some 
category that is being referred to by ‘we’, is to consider what may be a same 
use of the term in this segment - where there are quite different uses of it, 
clearly. That is, there are some uses in the middle which perhaps refer to a 
group whose incumbents are two persons. But the last use of ‘we’ in this 
segment, in the sentence “We got company,” looks to me, anyway, like a 
similar use to this first use. And now we’ve got an alternative category to the 
unknown category we’re dealing with, which is ‘company.’ Whoever ‘we’ are, 
‘company’ is something in alternation to that. And ‘company’ is a kind of 
interesting term. First of all, it doesn’t discriminate either as to single or plural 
personnel, and it is clearly a category that is in alternation to some other 
category. Now it’s interesting, for one, that it’s a category in alternation to 
some other category, but there are a variety of uses of ‘company’ which stand 
in alternation to a variety of categories. It is, in a way, a general alternation 
category. That is to say, this could be ‘family’ and that could be ‘company;’ 
this could be a variety of other things and that could still be ‘company.’ 

Lecture 4 
Tying Rules 

Last time I ended in the middle of a consideration of that first use of ‘we,’ and 
I made some issues about the kind of term it might be. I’m not going to 
continue with that now, because I’d rather hold off on considerations which 
are somewhat deeper than what I want to be doing, which is just going 
through some of the readily observable features of this segment. And when 
we’ve gone through those, we can deal with such matters as the ‘we.’ 

So let’s proceed now by first considering some of the ways beyond what 
we’ve so far gotten, that the utterances of this segment are tied together. In 
doing that we will of course - ‘of course’ because that’s the way the data 
happens to run — be re-observing things that we’ve observed earlier in the 
course. So let’s get at some of the ways that the parts are tied together - where 
that they are tied together is a fact, and would be part of the warrant for 
saying in the first instance that there’s a conversation going on, and perhaps 
even that what I picked out is an isolatable segment of it, i.e., that we have 
some way of deciding that this is ‘a segment of conversation.’ 

When I earlier introduced some sequencing rules I said about them that 
they could operate within a sentence, across sentences, and indeed across 
utterances. And we have a variety of them so operating here. We can stick 
initially to those that operate across utterances. They tend heavily to operate 
across consecutive utterances; that is, between two utterances, one of which 
follows the other. 

Now right off is what looks to be an instance of this Verb Followed by 
Pro-verb Rule: 

Joe : I still say though that if you take, if you take uh a big fancy car out 

on the road and you’re hotroddin’ around you’re you’re bound to 
get, you’re bound to get caught and you’re bound to get shafted. 
We- look 

Mel : Now did you do it right. 

In this instance the rule would deal with “are hotroddin’,” Mel’s “do” being 
the Pro-verb replacing, standing for, substituting for, referring back to Joe’s 
“are hotroddin’.” 

There’s at least one, and surely more, Noun Followed by Pro-noun: 

Mel : Not many people get picked up in a Pontiac station wagon. 

Joe : Now I agree it looks like a daddy’s- It looks like a damn 

mommy’s car. 


Lecture 4 


That is, “Pontiac station wagon” in one utterance, “it” in the next utterance. 

The character of these as rules in these cases will have to be examined, and 
I’ll try to say something about that in a while. 

One that recurs several times, and that’s rather more complicated, we 
could start out by examining in the sequence: 

Henry : And we also decided you’re a chicken shit. 

Joe : I decided that years ago. 

The term I’m interested in there is Joe’s “that.” There are several others like 
it. In Henry’s first remark of the sequence he says “That’s the problem with 
society,” and “that” presumably refers back to the long if-then clause of 
Joe’s, “if you take a big fancy car out on the road. . .” etc., etc. And there’s 
one of the same sort, internal to a particular utterance, Mel’s “Now did you 
do it right. That’s the challenge.” 

I stated initially that ‘that’ is a more complicated problem. That is so 
because, for one, ‘that’ can have a variety of other uses, and does, in the very 
conversation we have here. For example, Joe’s “I still say though that if you 
take. . .” right at the beginning of the segment, is a rather different thing. 
What that involves is that we have to have some way of locating the class of 
that,’ apart from the sheer occurrence of ‘that,’ which is possibly relevant to 
this rule. Where, let’s say for now that the rule works something like: 
Predicate ■*— ‘That’. 

The question of how we go about locating those ‘that’s which are relevant 
to this rule, will involve us in considering the character of all of these rules 
when they’re operating across utterances. And let’s do that, briefly, anyway. 
The occurrence of these rules, being not located within a single utterance, 
poses for us a problem: What form are these rules supposed to have, insofar 
as they are somehow regulating or guiding the Members’ actions; and what 
kind of procedures for analysis do they pose? 

At least a rough version of the problem might be as follows: The fact that 
there are these rules, and that they have a clear locus within a single sentence, 
say, operates to provide a maxim which is what we can call a Second Speaker’s 
Maxim, though ‘second speaker’ has to be given what may be seen as a special 
formulation. Somebody who wants to make themselves a second speaker to 
some utterance can have as a maxim for him that if he wants to tie his 
utterance to that of a preceding speaker’s, he can formulate some part of his 
utterance as a second part of one of these rules, if he can use some second part 
which can be tied to some then-made ‘first part’ of a then-made ‘first 
speaker’s’ utterance. That is to say, somebody who wants to make 
themselves a second speaker, can make somebody a first speaker. It can be 
anywhere in the conversation. He makes some part of what some speaker 
did, a ‘first part’ for one of these rules. He doesn’t start out, we take it, 
with a set of marked objects: First Parts for these rules. But by his speech 
he makes some part of what some speaker did, a ‘first part’ for one of the 


Part II 

So the analysis starts out with the locations of the second parts of these 
rules. And it’s in that sense that someone formulates himself as a ‘second 
speaker,’ i.e., the maxim doesn’t govern first speaker’s actions. What we’ve 
got then, is something like this: There are these rules which may have an 
unspecified environment of use, which could be a single sentence, some 
sequence of sentences by a single utterer, or the utterances of several persons. 
If they’re going to be used by some group of speakers, then it’s the business 
of somebody who is formulating himself as a second speaker to invoke them 
and provide for the relevance of a first item by using something that is clearly 
a second item. When someone does such a thing we get, quite obviously, a 
sense in which he ties his utterance to that of some preceding speaker. 

Let me indicate that it’s perfectly possible to set oneself up as a first speaker 
by the use of other sorts of rules. For example, if you ask a question, then the 
fact of a question sets one’s position as ‘first speaker’ to which the rule Q — * 
A, Question Followed by Answer, is locatable by the fact that there’s an 
observable ‘first occurrence.’ And I think it is the case that there are a variety 
of ways that persons can formulate themselves as first speakers for even some 
of these Second Speaker Rules, though that’s a rather more subtle problem 
and we’ll deal with it as we come to handle data in which it’s present. 

And I want to introduce an unspecified caveat to the Second Speaker’s 
Maxim: It is the case that persons can in one way or another move to prevent 
the use of their utterance as something to which another can tie an utterance, 
and they can move to encourage it. So while the maxim we set up can operate 
independently of what the formulated ‘first speaker’ does, there are things a 
speaker can do to effect what treatment their utterance may get. That there 
are, is something that is for at least a certain kind of conversation, quite 

But for utterances handled by the Second Speaker Rules, we want to start 
out with the second item and find a way of finding the first. And we have to 
have some clear way of deciding that we have a second item. Some of them 
may be quite clear as, at least, possible second items. The various pro-terms 
might be quite clear, but we take it that ‘that’ is not clear. How then do we 
find out whether we have a ‘that’ which ought to be seen as a possible second 
item, so as to start the search for a first, so as to see, then, whether that person 
has used the rules to tie his piece of talk to a preceding piece of talk? What 
we’re considering then, is the sequence that the analysis has to go through, 
and what is has to include. 

Apparently what it has to include is an analysis of the syntax of the 
sentence. And that’s important. It’s in the first instance, I think, the syntax of 
the sentence that tells you whether you have such a ‘that’ as is a likely second 
item. The differences of the ‘that’s are the different syntactic functions that 
they serve. Those that we’re dealing with here as second items are probably 
easier seen as the objects of the sentences, and those that are not are things 
like, particularly, conjunctions. Now those sentences in which the relevant 
‘that’s occur, have them in a very clear position. “That’s the problem with 
society.’’ And what’s invoked by “that’’ is the predicate of the preceding 

Lecture 4 153 

utterance; that is, the problem of society is that if you go out and hotrod, etc., 
certain things will happen to you. 

And for “I decided that years ago,” ‘I decided years ago’ is an adverbial 
phrase, I suppose; ‘that,’ the object, replaced with ‘I’m a chicken shit.’ And 
I take it that that at least holds for the material that we have. I also take it 
that you oughtn’t to get terribly frightened, since you could easily enough see 
what the object of the sentence is, as compared to ‘that’ being used as a 
conjunction in, for example, ‘‘I still say though that if you take ...” 

But the important thing is that to locate those second items we have to go 
through some crude or not syntactic considerations, and it’s not the word’s 
occurrence that does that for us. If we then locate a ‘that’ that is an object of 
a sentence, we have the basis for then proceeding to consider whether there is, 
and what is, that thing to which it’s tied. (In the materials we have, they tend 
to be whole predicates, I guess, of preceding utterances.) and if not whole 
predicates, then pretty near whole predicates.) Now that’s something that we 
haven’t previously considered. That is, we have not previously found 
situations where syntactical analysis was required before we could do the 
analysis we needed to do. 

There are a variety of other things which are, in the case of this 
conversation, perfectly simple to see. Such things as the simple use in some 
second utterance of the same subject or the same subject plus verb or the same 
verb, as is used in a previous utterance. There are a bunch of such things; for 
example, the use of ‘you’ in Joe’s ‘‘if you take a big fancy car out on the road 
and you’re hotroddin’ around you’re bound to get caught,” followed by Mel’s 
“You wanna try and do it right so you do not get caught.” That ‘you’ is an 
exceptionally important phenomenon. It’s a definite pronoun, and it has no 
special reference at all. It certainly doesn’t mean you, the listener. I’ll consider 
that sort of ‘you’ when I get finished dealing with these rather surface aspects 
of the way the conversation is tied. 

Verb phrases recur in first and second utterances; there’s “picked up” 

. . .“picked up” in some sequence, ‘tear down”. . .“tear down ” in another, 1 
and others. 

Now there are some uses of the rules which start out by having the whole 

1 The reference here is probably to the following: 

Joe : and the guy will pick the guy up in the dirty grubby tee shirt before I 

before he’ll pick me up. Just- just for- 

Mel : Not many people get picked up in a Pontiac station wagon. 


Henry : Hey don’t tear him down 

Joe : I’ve been torn down for- 

Mel : Ok 

Henry : We got company 


Part II 

rule introduced in a single utterance, and then, by virtue of the fact that the 
rules are repetitively operatable - as we said earlier, you could have N ■*— P 
•*— P, indefinitely - what we get is: N •*— P in one utterance, and then the 
continuation in the second. So we have, “Joe, face it you’re a poor little rich 
kid,” then several lines down there’s an elaborate way of relocating that 
subject, and a continuation of the use of ‘you’ as the subject thereafter, quite 
clearly tying that to the first use. 

Henry : Joe, face it, You’re a poor little rich kid. 

Mel : Okay Now you’ve got that out of your system. Now you’re a 

poor little rich kid we’ve told you that. 

Henry : And we also decided you’re a chicken shit. 

Mel : Now let’s see what else can we decide about you? 

I’ll consider this one, also, in more detail, because it’s not a perfectly simple 

There’s a rather interesting one involved in the following sequence. We get 
“Bonneville” •*— “it,” i.e., N ■*— P, and then “Pontiac station wagon” 
“it.” They’re talking about the same car, and it’s at least plausible that the 
reference of ‘Pontiac station wagon’ is to ‘it’ in the preceding utterance, as its 

Joe : In that Bonneville of mine? I could take it out with me . . . and the 

guy will pick the guy up in the dirty grubby tee shirt before I before 
he’ll pick me up. Just- just for— 

Mel : Not many people get picked up in a Pontiac station wagon. 

Joe : Now I agree it looks like a daddy’s- It looks like a damn 

mommy’s car. 

I said it’s interesting because there’s a tremendous economy in this conver- 
sation, and in many such conversations, involving the use of nouns - they’re 
just not much used. And when you get one used, it’s doing some work. The 
compactness of these conversations is handled in part via the tremendous use 
of these pro-terms, which just make up the bulk of talk. That can be seen at 
first glance by virtue of the fact that there’s only one use of a personal name 
in the conversation, and that use has a major significance, it marks a big 
change in topic. 

In the case of “Bonneville”. . .“it,” “Pontiac station wagon”. . .“it,” the 
renaming of the car is clearly a recharacterization of the car. I say ‘clearly,’ in 
the sense that if you knew it was the same car, then you’d take it that 
something different is being pointed out about it. The person who originally 
used the term ‘Bonneville’ acknowledges that something different has been 
done (Joe, with his “Now I agree it looks like a . . . damn mommy’s car”), 

Lecture 4 


and some people who have seen the conversation and didn’t know it was the 
same car, took it that in fact an attack was being made, without regard to 
whether ‘Bonneville’ and ‘Pontiac station wagon’ were different names of the 
same object. The fact that that’s done by change in name would suggest that 
whatever generic critique one could make of Russell’s notion that names are 
disguised descriptions, nonetheless in this sort of situation that’s the way the 
names are being used. So that ‘Pontiac station wagon’ is a transformation of 
the name of the object being initially referred to, which transforms the 
character of that object from something which, let’s say roughly at this point, 
could be a thing which you could approximate hotrodding in, to something 
which you couldn’t possibly approximate hotrodding in. Why you couldn’t 
possibly, is something we’ll consider eventually. 

Let me just note about the range of sequencing rules, that they can be used 
just as sheer indicators that there’s a conversation going on. And one way you 
can characterize people who can’t converse is that, so far as you can tell, you 
could perfectly well disorder the conversation without having any effect on it. 
That is, you could juxtapose any of the parts. Given the way these things 
operate, if we have these rules tying pieces of conversation to other pieces, you 
can’t disorder them. You can’t put an utterance that contains the second for 
some particular first, ahead of it. It matters what sequence they come out in. 
And people make it their business to make it matter. 

There are also some much more interesting, perhaps less obvious, and in 
any case less overt and ‘technical’ - in the sense of having to do with the 
names of these words - ways in which this conversation is tied together. But 
to get at them I’ll have to do a whole big analysis and I want to deal with 
things that are right on the face of the conversation before we move on, so you 
can see first how obviously locked together it is. Therefore, I’m going to move 
to some other surface features of the conversation before dealing with the 
more elaborate ways it’s tied together. In doing that. I’ll be doing something 
which I ought to have done previously, and which may give some sop to 
people who are interested in personality . 2 And that is, there are some really 
quite striking ways that, even in this one page, these three guys speak 
differently. And these are quite obvious and observable things, though some 
of their technical features may not be obvious. For example, this fellow Joe 
recurrently uses a pattern like “you’re you’re bound to get you’re bound to get 
caught.’’ There are three or four such uses. It’s not an unheard-of thing. And 
it’s important to see that it’s not an unheard-of thing; that it is something that 
has been observed, and observed as characteristic of something. This is not to 
say that if it is characteristic of something, then he is one of them. But it is 
characteristic of the speech of very young children. It’s something called 
‘buildups.’ For young children, it may have as its basis the fact that they have 
a very limited grasp of grammar, and they literally assemble a sentence by 
putting its parts together, piece by piece. If you look at a book I’ve mentioned 
before, Language in the Crib by Ruth Weir, she has a discussion of these 

2 The title of the course was ‘Culture and Personality.’ 


Part II 

things (pages 81-82), and a bunch of examples. The kid there is, I think, 
something around two. But any of you who have heard young children talk, 
will have heard that sort of thing. It’s obviously not simply a stutter. And it’s 
extremely characteristic of this fellow, Joe. 

Another thing that is sort of striking occurs twice in Mel’s speech, and I 
don’t think anywhere else. It’s not a unique usage at all, and it’s perfectly well 
examinable as to what it can do. And that is, he uses the syntax of questions 
without making - for utterance-rule purposes - questions. That is, he builds 
a syntactical question without intending that at the end of it he would stop 
talking and somebody would answer. He does that in quite obvious ways by 
using the reversal pattern of questions: “Now did you do it right,” for 
.example, and “Now let’s see what else can we decide about you.” That is, 
“. . . can we . . . ,” “. . . did you . . .” are such kinds of things. And I’ll 
consider those, also. 

The third fellow, Henry, has what looks like an exceptionally idiomatized 
way of talking. Three of his whole utterances are single pre-constructed 
sentences, akin to proverbs in a way. That is, they are sentences that he didn’t 
have to construct for this occasion, that can by used by whomsoever, that 
come on just as that - a sentence. Things like, “That’s the problem with 
society,” “face the music,” and “we got company.” 

Lecture 5 

Tying rules; Insult sequences 

I’ve been examining how the parts of this segment of conversation are tied 
together, dealing first with some very small ‘pair-ties.’ We’ll move on to 
much more elaborate ones as we go along, and then get into what this 
conversation is about, and how it works with respect to that kind of matter. 
The kinds of rules I’ve been offering are by and large not at all constrained by 
what the conversation happens to be dealing with. Eventually we’ll see that 
there are no untied utterances in the segment. And by that I mean, no 
utterance for which there is not at least a single tie. We’ll find that for some 
there are far more than a single tie. I haven’t worked out a measuring system 
to compare conversations in this regard, but I take it that that might be 
doable in any event. 

Let me go over some of the rules I’ve already given. We have, for example, 
some purely-internal-to-a-sentence rules, like the one for making an indepen- 
dent clause out of what might have been a sentence, and for making your own 
utterance a dependent clause in that sentence, and for making your own 
utterance a dependent clause in that sentence. It runs something to the effect 
of: If some given utterance is ended by a noun, then participialize that noun 
and use it to begin your own utterance. And of course, besides participializing 
that noun you have to be otherwise syntactically consistent with the preceding 
utterance. Just consider: “We were having dinner”. . .“dining on roast beef.” 
You can make a whole bunch of them in just that way, although that’s not 
the only way you make dependent clauses. 

The second rule concerns prepositional phrases. They consist of a preposi- 
tion plus a noun phrase. If an utterance ends with a preposition, then if you 
begin the next with a noun phrase that could be the noun phrase for that 
preposition, you tie your utterance to it. 

A third rule concerns the use of conjunctions. Now the conjunction is rather 
more complicated because the syntactical consistency is much more impor- 
tant. If an utterance ends with what might be a sentence, if you begin the next 
with a conjunction and otherwise make it syntactically consistent, then again 
you can tie the second to the first. 

Then there are a variety of rules which, while they could operate within a 
sentence or within a single utterance composed of several sentences, they can 
also operate across utterances. Given that for these that I’m going to list, we 
start out with the second item for the rule and not the first, it might be 
convenient not to write them as, e.g., N P (N followed by P), but as P — ► 
N, so that we know that we’re starting with P. So there’s that one for 
pronouns and nouns, there’s the same sort for pro-verbs and verbs, there’s the 



Part II 

one for ‘that’ (with a special constraint on syntax) and predicates; now there 
are some more. 

Some of the others, once we mention them, will permit us to see a rather 
clear difference between those things which we identify in the first instance 
from what turns out to be the second object, and those we identify by noticing 
the first object. For example, we can look at a question-answer sequence, 
which is as clear a version as you want of starting out with the first part. In 
this conversation you get things like: “What school did you go to?’’ I went to 
Palisades.’’ and “You’ve been kicked out of school once.” “Yeah.” Now for 
those, it’s by virtue of the fact that the first item provides that the next is 
going to be a second, that those seconds are clearly seconds. They are not 
otherwise so identifiable. They’re not even identifiable as answers to questions. 
But there’s no doubt, I take it, for anybody, that things like “Yeah” in that 
sequence, are second items. But what we have in these other kinds of rules, 
is that the first items are not identifiable as ‘first items.’ It’s by virtue of the 
fact that the second is clearly ‘a second to something’ that we go back and find 
a first, as compared to, say, the question-answer sequence where we identify 
the first. That means, for one, that you get a kind of distribution of 
compactness in the question-answer sequences. Given the fact of the question 
providing the relevant sense of the thing that ought to follow it, nothing 
much has to go into the answer. And a zillion questions could be answered by 
‘yes’ or ‘no.’ 

Now we can find some instances of this question-answer sort of rule in this 
piece of conversation. There is at least one command and then return-to- 
command - it could be either acceptance or rejection of the command, but it 
would be clearly seen as one or the other. For example, ‘Joe, face it . . . ,” 
“Yes . . .” (I’m not saying the ‘yes’ is an acceptance. I’ll analyze it eventually. 
It’s a much more careful thing than an acceptance in that case.) It is by virtue 
of the command being seen as the first item, providing for a relevant second, 
that the seconds have their character, though some answers are obviously 
answers - not many. 

There’s one which you can play conservative or non-conservative on, as to 
how you write the rule, but let’s say: Incomplete Utterance followed by 
Interrupter Term. I say you can play this conservative because you might not 
take it that the utterance was clearly incomplete. And you might start out 
with the fact that it looks like you’ve got an interrupter term, and work 
backwards: interrupter term — ► incomplete utterance. 

I take it that that holds in its strong version - that is, incomplete utterance 
first - for the sequence: 

Mel : Now let’s see what else can we decide about you? 

Henry : Hey don’t tear him down 

‘Hey’ is clearly an interrupter term. You might not take it that it’s obvious 
that “Now let’s see what else can we decide about you” is an incomplete 
utterance. To get some feel for that, reconsider the earlier statement that is 

Lecture 5 


syntactically alike to it, “Now did you do it right.” And I want to cross out 
the question mark at the end of “Now let’s see what else can we decide about 
you?” because in my last listening to that piece of conversation I find that I 
now think that there is no intonation at the end that leads one to feel it’s a 
question, though of course it has that reversal construction of a question, i.e., 
not “let’s see what else we can decide ...” but “let’s see what else can we 
decide ...” 

There is a class of tying verbs that are used. There are a lot of verbs that 
may tie, but there are some verbs which, when used, are just tying verbs. Like 
‘agree.’ When somebody says “I agree,” they’re tying that statement to some 
other statement of somebody else’s. So you could write it as some utterance 
•«— tying verb, but it would probably be done as tying verb — *• some 

Now there are some things which can occur within a single utterance or 
could even occur in single sentences, which also can occur across utterances. 
When they occur across utterances they are clearly serving to tie utterances 
together. And those are what we could call lister terms. The most obvious 
lister terms are, of course, things like the ordinal numbers: First, second, third, 
as we get in the next segment of this conversation, in the series starting with 
“Well first of all you must be crazy . . .” But there are some used in this 
second segment to do that kind of work, that are not the ordinal numbers. A 
series can be explicitly started with a first, L, — ► L 2 . Or you can have a first 
occurrence of a lister that is making itself follow something else, where the 
other has not been specifically introduced as the first of a list. In this segment 
Mel says “Now you’re a poor little rich kid, we’ve told you that.” You might 
say that ‘now’ is a lister term. It’s not clearly one, it has plenty of other uses, 
and you’d have to make a good argument for it. But it’s followed by “And 
we also decided ...” which turns it into a ‘first’ where it could have been just 
a single. 

A lot of other rules can be constructed, but with these you can certainly get 
some idea that these people were in some way or other talking to each other, 
and making the fact that they were talking to each other a big part of their 

Now, what I’m going to be doing is taking small parts of a thing and 
building out from them, because small parts can be identified and worked on 
without regard to the larger thing they’re part of. And they can work in a 
variety of larger parts than the one they happen to be working in. I don’t do 
that just as a matter of simplicity, but as I mentioned earlier in the course, the 
image I have is of this machinery, where you would have some standardized 
gadget that you can stick in here and there and that can work in a variety of 
different machines. And you go through the warehouse picking them up to 
build some given thing you want to build. So these smaller components are 
first to be identified because they are components perhaps for lots of other 
tasks than the one they’re used in. But then they’ll be fitted together into some 
actual single larger component in this case. 

Okay, let’s look at this one for a while: 


Part II 

Henry : Joe, face it, You’re a poor little rich kid. 

Joe : Yes Mommy. 

Henry : Face the music 

We could start off simply enough by saying that it is a recurring part of, let’s 
say at least, teenager’s conversations, that an insult is properly followed by 
another, or a return to it. 

In fact, that little thing itself is the basic component of some very 
standardized teenagers’ games. They are alternation games, they have names, 
and one of the names is in fact used later on in the conversation. There is a 
considerable literature which pretty much first emerged into literature in 
a paper by John Dollard in 1957 called ‘On playing the dozens,’ which is a 
Negro kids’ game, and it’s an insult game. The kids engage sequentially in 
insulting each other. They tend to do it before some audience, where the 
audience, by its reaction, decides that the game is over. In that form it’s 
been well written-up in a variety of papers. Dollard’s paper is in a book by 
Roger Abraham on that sort of phenomenon, called Down in the Jungle 
( 1964 ). 

The game was apparently not much thought to be one that white high 
school kids play, but they do. And one of the names in that environment is 
‘tearing down,’ which is what the action eventually gets called in this 
sequence. It is not, however, to be seen as getting its analysis by some simple 
historical expansion - which is one way that people have considered it - that 
is, that it passed from Negro kids to white kids at some time. Because some 
of its features, like this ‘Mommy’ bit, are fabulously old. 

The ‘Mommy’ return is extremely classic in such kinds of things. I’ll give 
you a quotation. This guy Eric Partridge, who writes all these things up, has 
a book called The Shaggy Dog Story, and in it he reports what he calls an 
ancient Greek story. 

A pert youth meeting an old woman driving a herd, called “Good 

morning mother of asses.” “Good morning, my son,” she returned. 

Apparently that’s one of the classic ways that you handle an insult, i.e., if it’s 
an insult for which the most elegant return would be to make my status a 
consequence of yours. And kinship obviously is the most powerful way of 
doing that, like “You’re an ass.” “Thanks Dad.” Those kind of things are 
obviously extremely powerful and by no means new, by no means to be 
accounted for by the fact that in Detroit the racial barriers are broken down 
so that Negro kids and white kids are closer to each other - except 

There are several reasons I’m going through the fact of this game. One is 
that by treating the first event as possibly the first event in such a game, by 
doing a second, that’s a way of giving a special characterization to the first 
insult as something that is not serious and is not going to be heard as serious. 
It also sets up a challenge, which is, Okay you started, I came back, go ahead 

Lecture 5 


if you’re willing to. And Henry withdraws. That is, the move that would 
accept the sequence as a game and continue it, would be for him, since he has 
a slot, to come back with another insult. But he closes it off with “Face the 
music.’’ It’s quite apparent that what he’s done is to withdraw, in that when 
it rapidly turns out that he has an ally in that thing, it starts up again, as you 
can easily enough see. 

Henry : Joe, face it. You’re a poor little rich kid. 

Joe : Yes Mommy. 

Henry : Face the music 

Mel : Ok Now you’ve got that out of your system. Now you’re 

a poor little rich kid we’ve told you that. 

Henry : And we also decided you’re a chicken shit. 

While we’re on this sequence, we might as well note that one of the things 
we earlier pointed out is also present here. I proposed that if it were the case 
that we had the form of pronoun-noun relationship where the pronoun 
actually precedes the noun and is not the second that’s discovered first, but is 
first, then something special is going on. And the ‘it’ in “Joe, face it’’ is such 
a one. One pretty clear thing it’s doing is, it keeps that utterance open to then 
flip out the item that completes the N-P pair. 

There’s a whole bunch of ways that statements that might be closed by 
virtue of the fact that they’re sentences, are kept open in this first segment. I 
pointed to some of them already, though I haven’t worked out how they 
work. For example, those special kinds of questions that this fellow Mel uses. 
Another is that even if the game itself doesn’t go on, and one is restricting 
oneself to an insult-and-retort sequence as one’s expectation, then apparently 
one of the things about the insult-retort sequence (and this is another 
tradition which is perhaps independent of or pre-existent to this game) is 
that there’s a big crucial thing on having the last word, for some reason or 
another. And by first producing “Face it’’ and saving the formula “Face the 
music,’’ then, whatever the retort is, Henry has the last word, because the 
formula is tied to his first usage, and he can then fit it in and close the sequence 
off. So, where you have a formula that you’re invoking by splitting its usage, 
that may be a significant thing to do, and in this case it clearly is. And 
there’s plenty of literature about having the last word. There’s a classic 
thing to look at, an absolutely fabulous book I mentioned before, The Lore 
and Language of School Children by Peter and Iona Opie (Oxford University 

Now I want to deal with another little one, one which what I have to say 
is somewhat beyond my level of believability - which I guess is much more 
extravagant than other people’s - but nonetheless it’s possible. And that’s the 

Henry : And we also decided you’re a chicken shit. 

Joe : I decided that years ago. Hell with you 


Part II 

The question is, what kind of a return is “I decided that years ago, hell with 
you.” Clearly ‘‘hell with you” is of some relevance. It’s an indication of 
courage, or something. Now it’s perfectly clear that “I decided that years 
ago” is not an agreement. It’s instead, a rejection. Why should it be a 
rejection? It looks like a perfectly good agreement, nonetheless it’s perfectly 
clearly not an agreement but a rejection. There is an obvious enough 
formulation, except it’s as I say, beyond my belief that they could be that 
ingenious, and that is that what he’s doing is to set up an antinomy. And he’s 
using the fact that he’s setting up an antinomy to undercut the statement. An 
antinomy is a statement that, if it’s true, it’s false, and if it’s false, it’s true. 
The classical one is “I’m a liar. Everything I say is false.” Here, if it takes 
courage to recognize that you’re a coward, then to have recognized that you’re 
a coward took courage, and therefore you couldn’t be a coward. 

Given such a possibility, I would then try to see, is it the case that people 
in their ordinary affairs do use antinomies. And I haven’t really looked very 
hard so far, but I found at least one, and it’s a classic one, also. It comes from 
this book, Style in Language, edited by Thomas Sebeok, a paper called ‘Oral 
styles of American folk narrators’ by this American folklorist, Dorsen, on 
page 41. He’s reporting some story he was told: 

The next supposedly true happening, where Art is asked to tell a lie and 
says he has no time because so-and-so has just had an accident and he 
must go to a doctor - which is a lie - is an international folk tale 
attached to various American yarnspinners 

Lecture 6 


I’ve been talking about this bunch of tying rules. There are a lot more, even 
of the sort that we’ve been laying out, apart from any that are more intricate 
than those, but that’s enough for now. The question is, so there they are, why 
bother talking about them? What is it that they do? To some extent it’s 
terribly simple to see. It’s one sort of task for the hearer of some statement to 
determine what it is that the person who produces some noun is referring to, 
if he’s referring to anything: What’s an ‘automobile discussion’? What’s a 
‘car’? etc.; it must mean this or that. That’s one sort of task. When the tying 
rules are used, another sort of task is imposed on the hearer: To decide what 
it is that the tied term - for example, ‘it’ - refers to, requires finding 
somewhere in the conversation that the term it ties to occurs. 

So in the use of these, there’s then a required piece of work for the hearer, 
which involves collaboration in making out the conversation. Understanding 
the term implicates the hearer in the conversation - and of course provides 
that in the very use of such an object the speaker has implicated himself in the 
course of the conversation. The use of these, then, is not to be seen as simply 
a way of, for example, avoiding redundancy or making variety or whatever 
else, but it provides an order of work - and is produced by an order of 
work - which is at least to some extent different than the work involved in 
using a name, etc. So that, at least in the first instance, is what the use of these 
tying rules involves. And that’s then, the way in which they make for the 
sense that the various participants have, that a conversation is taking place; 
where the participants are implicated, and they are involved in working at all 
of its parts, as a set of parts. 

Now the work that’s involved can be sort of minor in that someone may 
claim, or may in fact not be able to see what it is that some pro-term refers 
to, i.e., what term it refers to. But it can also be extremely important. It can 
be the case that some usages provide that if their hearer wants to play dumb, 
then he’s going to have an extremely touchy task to do that. That is, the 
routine use of these things may operate to provide a great deal of information 
for the one who uses them; for example, that on some occasion the other does 
in fact understand what one is talking about, by virtue of the fact that he can 
continue to put in terms where, quote, nothing explicit has been said - or not 
much explicit has been said. Eventually I’ll give you some instances of usages 
which are strategically very powerful in the sense of forcing someone to 
acknowledge that they do in fact understand something, where the fact that 
they do understand it has tremendous consequences. But for now, I just 
wanted to raise it. The initial remarks about tying rules would be enough to 



Part II 

give them some strong warrant. The things I will mention eventually are just 
elaborate technical developments on them, and they’re of some interest but 
they’re not that enormously core. 

Let’s go back, then, to dealing with bits and pieces. In the first pair of 
utterances after the introductory sentence of the second segment, we have 
‘you’ used as basically the subject, i.e., “. . .if you take a big fancy car out on 
the road and you’re hotrodding around you’re bound to get caught and 
you’re bound to get shafted.” “Now did you do it right. That’s the challenge. 
You want to try and do it right so you do not get caught.” I want to make 
some remarks about ‘you.’ 

First let me list some references to the extremely large literature on terms 
like ‘you;’ terms such as I, you, he, this, here, now, it, and the like. They’ve 
been subject to a lot of study by philosophers and logicians, and if one is 
interested in conversation, one might well look at that literature to see what’s 
been done on the phenomenon of conversation. Some of the literature is at 
least historically interesting. There’s chapter 7 of Russell’s book, Inquiry into 
the Meaning of Truth, and he calls them ‘egocentric particulars.’ Then there’s 
Nelson Goodman’s book, The Structure of Appearance, chapter 11. He calls 
them ‘indicator terms.’ Then there’s Reichenbach’s book called Elements of 
Symbolic Logic. He has a long chapter on the analysis of conversational 
language, which isn’t what it says it is; not at all. It’s important to see that it’s 
not - and to see that nonetheless it’s pretty unique in saying that it is, anyway. 
That’s chapter 7, and he calls them ‘reflexive terms.’ Then there’s Quine, 
Word and Object, section 21. He also calls them ‘indicator terms.’ Then there’s 
a lot of articles in the philosophical journals. I’m not going to cite them all. 
Some of them use the names I’ve given you, but one that’s relevant in that 
it uses its own name is a paper by Bar Hillel in Mind, 63 (1954), titled ‘On 
indexical expression.’ That’s what he calls it. Then there’s an old monograph 
by a linguist named Cullenson, called ‘Indicators,’ Language Monographs, 17 

The core thing about these terms, and why they’re of interest to the 
logicians, is that they have an extraordinary transiency of reference. So the 
issue is, if you’re dealing with ‘the truth or falsity of propositions,’ then how 
are you to interpret a proposition which has the term ‘now’ in it, or ‘he’ in it, 
or ‘you’ in it, etc. Much of the work has been directed to trying to find a way 
of rewriting statements or sentences that have those terms in them, so as to 
provide for the fact that they could be true or false. I won’t be considering 
such issues. 

But there’s one immediate consequence for my previous discussion. One of 
the things you ought to see is that if any such terms occur in immediately 
juxtaposed utterances, that would not be necessarily at all an instance of 
something that you would want to call a tying-rule case. My proposal that ‘If 
the same subject appears in two consecutive utterances then those utterances 
are tied,’ would not at all obviously be the case if ‘you,’ and then ‘you’ 
appeared. For one, if there are two different speakers, they’re not at all 
referring in any obvious way to the same person. In fact, if you’re going to use 

Lecture 6 


these in focusing on the tying operation, what you would want is that 
whichever term appears in the first utterance, the alternative term appears in 
the second. You could have an instance of tying if you had ‘y° u ’ ‘I,’ or T 

— *• ‘you,’ where ‘you’ — ► ‘you’ is not at all an obvious such case. So these 
terms pose a problem for working on the tying operation, but it may be 
worked out. In any event, the usual ‘you’ — *■ ‘you’ problem is not present in 
the example we’ll be considering. 

In this case, “. . . if you’re hotrodding you’re bound to get caught . . . ,” 
“You want to try and do it right so you do not get caught,’’ we have what 
is a very recurrent use of ‘you,’ and it’s what, if Reichenbach were to talk 
about it, he would immediately translate as ‘one,’ ‘someone,’ or something 
like that. It is not referring to the person being addressed. It’s what we could 
call, to start off with, an ‘indefinite use.’ And one of the things we want to ask 
is, how does it come about that ‘you’ is used to make that indefinite 
statement? Why don’t they say ‘one’? Are they equal? 

I think there are more than a couple of things involved, and I’ll try to 
mention most of them that I know of. Let me first discriminate a couple of 
possible senses of the term ‘ambiguity.’ One characteristic way it would be 
used is in an either-or relationship - either you mean this or you mean that, 
it’s not clear which, and they don’t have anything much to do with each other. 
Another usage of the term - and that usage is most considered when you’re 
dealing with poetic language - is this-and-that. For poetic language, one 
question is, how do you go about building richness into a compact form. One 
way is to get words which mean this and that, whatever this and that may 
mean. And the usage of ‘you’ that we’ll be considering is, I think, a case of 
the latter. ‘You’ is a very good term for attempting to build ambiguity of the 
this-and-that sort. 

I think the analysis ought to run something like this: What we have to do 
is to try to construct what a procedure might be for determining what it is 
that’s being referred to when somebody says ‘you.’ Is there a procedure? If 
there is one, what does it look like? And what are the consequences of there 
being one? 

Now ‘you’ in English is, at least in the first instance, systematically 
ambiguous, in that it does not discriminate between singular and plural 
usage. And in the ambiguity of ‘you,’ one has the this-and-that format; that 
is to say, it is ‘you’ (you alone) or ‘you’ (you and others). When a person hears 
‘you,’ they then go through a procedure of deciding what it refers to. And 
that procedure has a first step, where if the first step seems fully adequate, that 
ends the procedure, but if it’s not, then one goes on. The first step in the 
procedure is the consideration of the applicability of the singular use, and if 
it’s the case that one comes up with, well maybe that, but certainly not only 
that, then one moves to the second step and engages in finding, now, some 
plurality for which the ‘you’ would be correct. And that plurality could be, for 
example as I mentioned earlier, some category which, once it’s found it would 
turn out that the singular is a member of. It’s also the case that one of the nice 
things about the non-differentiation and the plural character of you’ is that 


Part II 

there is no limitation essentially on the size of the population it can be 
referring to. 

The singular ‘you’ is used as simply a way of dealing with the persons 
present and being spoken to in an interaction. Plural ‘you’ is no longer 
controlled by that fact. It’s not at all necessarily the case that it’s only referring 
to the multiplicity of persons present and being spoken to. Indeed, as I 
mentioned earlier, ‘you,’ even in its singular case, is regularly a way of 
referring to that member of ‘they’ who happens to be present. “You 

That openness of the plural ‘you’ means that ‘you’ can in fact be a way of 
talking about everybody’ - and indeed, incidentally, of ‘me.’ But there’s at 
least a tactical difference between saying ‘we’ with that intention and ‘you’ 
with that intention. The difference is, ‘we’ has some plural reference, 
automatically. And it does not have, automatically, a plural reference which 
includes the person being spoken to. This is in English. In some other 
languages it’s the case that ‘we’ can only be properly used if it refers to [me 
and the person I’m talking to}. So you do not say ‘we’ when you’re referring 
to [me and my wife, and not you}. In English, we do. And ‘we’ can 
then - intentionally by the person speaking, or by the decision of the person 
being spoken to - exclude the person being spoken to. In contrast, if you use 
‘you,’ it at least includes the one you’re speaking to, and on their option or on 
your intention, insofar as those coincide, it can refer to anybody else, or to 
some category which includes everybody else. 

And those differences are extremely carefully focused on by speakers. I 
have a lot of very subtle usages which turn on those differences; for example, 
a woman is asked “Why do you want to kill yourself?” and she says “Well, 
you just want to see if anybody cares.” Now that use of ‘you’ in this case 
surely refers to her, but refers to her as a member of ‘anybody,’ and thereby 
provides that it is only incidentally her reason, but it’s anybody’s reason, and 
thereby is not attackable as peculiar. It is offered as proverbially correct. 

Those kinds of uses are recurrent, and apparently quite powerful. I have 
one I just pulled out of the paper the other day, involving an interview 
between vice-president Humphrey and a Japanese newspaperman, on some 
relationship between Japan and the United States over control of atomic 
weapons, I think it was. And Humphrey is asked about America’s policy, and 
whether Japan would have more say. And, although he’s talking about what 
‘we’ will do, he formulates his remarks in terms of ‘you,’ e.g., “You want to 
give other people a chance,” etc. This involves now putting the Japanese who 
read it in the position of assessing his problem as though it were anyone’s 
problem in such a situation, and it’s nothing peculiar to America’s position 
that involves them in being hesitant; anybody in such a situation would be 

Now, one of the core things about this use of ‘you’ is that if in the first 
instance you find that the singular ‘you’ is inappropriate, you do not move to 
a sense of it which excludes yourself, but to a sense of it which is much larger 
but at least includes yourself. It may also, by interpretation, include the 

Lecture 6 


person who has spoken - as you see it, or as they see it, or obviously. But 
nonetheless, you are also included, which ‘one’ does not do. And ‘we’ does not 
do, and any of the other terms do not do - ‘he’ doesn’t, ‘they’ doesn’t, etc. 
But ‘you’ uniquely does. So it isn’t the case that ‘you’ and ‘one’ are 
equivalent. 1 And it does seem to be the case that the procedures for 
determining what it is that ‘you’ refers to — if the procedure is something like 
I make it out - is crucial to the special use the ‘you’ gets and that no other 
pronoun term gets. That is, ‘you’ is the term which gets used in specifying a 
proverb, or a proverbial type of frame. 

Returning to this sequence, there are other things involved in the use of 
‘you.’ One of them is another kind of ambiguity. We get this statement of 
Joe’s, “if you go hotrodding you’re bound to get caught.’’ If it’s some generic 
type of argument of an if-then sort (though there’s no explicit ‘then’), it could 
be delivered as, “if anyone goes hotrodding . . . ,” “if a kid goes hotrod- 
ding ...,’’ etc. What the ‘you’ does is, beside making it an argument, it 
borrows the form of a warning. That’s the way warnings get delivered. “If 
you do the following, then this will happen to you.’’ And the relationship, for 
kids anyway, of those generic if-then types with warnings, is probably 
something quite ambiguous. They first learn the warning version. They then 
have to learn to separate the pure argument type. And for some matters, I 
take it, it must be quite important that one never makes a complete break, 
and that there are ways of making the assertion not merely an argument, but 
also a warning, or a piece of advice. And clearly one way that would be done 
is to use the form of the warning, which involves the use of ‘you’ in that 
phrase. More generally, if you count the number of uses of ‘you’ in these 
proverbs, and their power - which is tremendous - you have a lot of 

Now I want to consider the sequence: 

Me/\ Ok Now you’ve got that out of your system. Now you’re a 
poor little rich kid we’ve told you that. 

Henry : And we also decided you’re a chicken shit. 

When we started out the discussion of this segment, I pointed out right at the 
beginning that we have here, again, as we have in the first sentence of the 
segment, a collaborative production of a sentence. In this case there are several 
ways that the parts are tied together. It’s not simply the ‘and,’ but the subject 
used in the first utterance is retained in the second, i.e., the ‘we’ - and that’s 
extremely crucial. And the relation is further established by the ‘also,’ which 
provides that it’s building on what has just been said. 

Simply by virtue of the set of properties we got, we might feel quite 
confident in saying that what we’re getting at this point is the forming-up of 

1 Several people in the class speak up with difficulties about the references of the 
terms, including a disagreement with Sacks’ proposals about what ‘one’ does and 
doesn’t do in contrast to ‘you.’ Sacks’ response is that “ ‘One’ doesn’t include the self 
or the other necessarily. [It] doesn’t include anybody necessarily.’’ 


Part II 

a second group within the group, which second group consists of as members 
the two participants to this sentence. Let’s look at it a little more - not that 
we’re going to find out that it’s wrong, but just to see how it works out. Until 
that point, we really haven’t had such a demonstration of the two-against-one 
characteristics of this conversation. When Joe got finished with his first long 
speech, the two others spoke essentially at the same time and took different 

you’re bound to get caught and you’re bound to get shafted. 
We- look 

Now did you do it right. That’s the challenge That’s the 
challenge You wanna try 
That’s the problem with society. Hahhh 
and do it right so you do not get caught. 

That’s the 

And when we got to that point where Henry starts to attack Joe, he did it a 
bit on his own, and closed it off - as I’ve suggested anyway. 

Henry : Joe, face it. You’re a poor little rich kid. 

Joe : Yes Mommy. 

Henry : Face the music 

Now Mel comes to the rescue, in a way. He ties his own statement back to 
the preceding attack on Joe, with the ‘now,’ and the reassertion of Henry’s 

Henry : Face the music. 

Mel : Ok Now you’ve got that out of your system. Now you’re a 

poor little rich kid we’ve told you that. 

He reasserts the statement as ‘we:’ “we’ve told you that.’’ That ‘we’ can have 
a strong and a weak usage. It could be strong if it were the case that what he’s 
saying is, ‘Henry told you that, and he is our representative.’ That is, the 
single person who spoke is a member of ‘we.’ That would be a chancy kind 
of thing to do, in two ways: One, Joe could say, “No, ‘we’ didn’t tell me that, 
he told me that.’’ And two, Henry could say, “No, ‘we’ didn’t tell him that, 

I told him that.’’ So he’s taking kind of a chance there, to go out on his own 
to establish what would be protective for the two of those fellows if it gets 
accepted. Given the possible chancyness of it, he does in fact establish it was 
weakly correct by virtue of the following: ‘We’ can equal, besides ‘he, as a 
representative of us,’ also, ‘he plus I.’ And he has in fact said it just now, and 
he saves the “we told you that” until he said it, so it’s weakly true at least, 
that ‘Henry said it and I said it,’ and that is ‘we.’ 

Joe : 


Henry : 

■ Henry : 

Lecture 7 

‘Hotrodders’ as a Revolutionary 

At the end of last time I was talking about that statement of Mel’s and most 
particularly the use of ‘we’ in it: 

Mel: Ok Now you’ve got that out of your system. Now you’re a poor 

little rich kid we’ve told you that. 

In it, at least as I see it, there’s two possibilities laid out. One is an offer made 
by Mel to Henry that the two of us are a group, of which what perhaps either 
of us say is said representatively for both. Alternatively, there’s weaker use of 
‘we’ which makes his statement at least safely true, i.e., where ‘we’ means in 
its weak sense, ‘you said it before and I said it now’, and thereby he needn’t 
be subject to any total rebuff. Now Henry takes him up on the strong use, and 
accepts the formulation in its strong sense, and in doing so he moves to a 
whole range of important doings. 

First of all, with his “And we also decided you’re a chicken shit,’’ he tacks 
on something that had initially been proposed by him prior to this 
eight-minute segment, as something that ‘we’ did, also. Since he was 
representative just a moment ago, he was also representative a while ago. In 
doing that, he also is involved in more sharply posing the fact that there has 
been a change in topic, as well as a change in the formation of the given 
group. The change in topic is that Joe is now clearly the topic. And Henry 
does that by indicating that a list has been started about Joe, with ‘also’ being 
this ‘lister term.’ And in doing so, he provides, then, for Mel what it is that 
he could do, which is of course to continue with the list. And that is an 
extremely safe way to proceed, given that in continuing he’ll be doing so as 
a member of this new subgroup within the group, which is in the first instance 
a strong one, since it contains at least a majority. There has been a first 
amalgamation within the group, with the three of them collaborating to 
produce a single sentence. Now there’s two against one. The two have joined, 
and the group has a distinct form. It’s not just a bunch of kids sitting around 
with no particular structure to the group. 

Once we have this majority situation, we get to what could be an extremely 
crucial point in it, which Henry proceeds to use in as strong a way as he likes, 
and that is, if there’s a majority in the group, then the majority can be 
representative of the group. If there’s a representative of the majority, then the 
representative of the majority can be the representative of the whole group, 



Part II 

and he who leads the majority is thereby the most powerful man in the group, 
in a special way: He’s not now just the most powerful man among the three 
but nevertheless subject to attack by either of the other two, but he’s the most 
powerful man within the majority, protected by the other in the majority, who 
would at least be concerned that their group last, and who might fight out the 
issue of who’s going to be the leader in that group, but be concerned to keep 
it that at least one of them will be the leader. 

What Henry rapidly does is, having got into a position where they’re 
clearly putting down Joe, he takes his first opportunity to now put down Mel. 

Mel : Now you’re a poor little rich kid we’ve told you that. 

Henry : And we also decided you’re a chicken shit. 

Joe : I decided that years ago. Hell with you 

Mel : Now let’s see what else can we decide about you 

Henry : Hey don’t tear him down 

Joe : I’ve been torn down for- 

Mel : ok 

Henry : We got company 

Mel : oh ok. 

Henry takes the opportunity to put down Mel by giving that command, 
“Hey don’t tear him down . . . We got company.’’ It is furthermore much 
stronger than just putting down Mel, because it also is, in the first instance, 
at least an attack on the new entrant. Bob, and also on a lot more, which I’ll 
go into. I lead up to this because what I want to do is focus on this term 
‘company,’ having just sketched the shifts that take place very rapidly right 
here, with Henry accepting the new formation and then moving to take that 
over, and thereby to take over the group itself. 

The category ‘company’ is, as I briefly mentioned earlier, rather a special 
one. And we ought ultimately to consider it in combination and contrast with 
a category like ‘hotrodders.’ Before dealing with the basic way it’s a special 
kind of category, I ought perhaps to mention some of its import, which can 
be seen without regard to what I’ll be saying. And that is, for them to 
formulate him as ‘company’ - and we’ve seen how you can say ‘them’ in a 
serious sense - is obviously for them to propose that whatever has been done 
by whatever institution that might tell this fellow Bob, or his family, that he 
has troubles, and whatever way this clinic that they go to might go about 
selecting personnel, and whatever position the therapist might be supposed to 
have in controlling the group, nevertheless he does not decide, and all those 
others do not decide, that Bob is a member. That’s for these kids to decide. 
For them, he’s company. And they seem to be able to do that. There is, then, 
some kind of independence in their determination of who’s a member, and if 
they’re controlling the conversation, then that determination on their part is 
indeed important. 

Now, holding that, let’s consider what kind of a category this category 
‘company’ is. I suppose about the most grandiose way I could choose - which 

Lecture 7 


is the way I choose - to bring the point home, is to quote something which 
has become, in its place, extremely famous and extremely fundamental. 
Though when you first hear it, unless you know about it, it’s not very likely 
that you’ll think there’s anything very special going on. It’s from chapter 14 
of Genesis. It goes like this: 

A fugitive brought the news to Abraham the Hebrew, who was 

camping at the terebinths of Mamre the Amorite, kinsman of Eshkel 

and Aner, these being confederates of Abraham. 

There’s a phrase there, that is, in the history of biblical criticism, one of the 
most fundamental of all in the bible. That is, ‘Abraham the Hebrew.’ There’s 
an immense literature on it, and here’s in part what its importance is: It is 
among the very few places in the bible that an Israelite is referred to by the 
term ‘the Hebrew.’ That term, ‘the Hebrew’ is only used by an Israelite for 
self-identification to a foreigner, or by foreigners about Israelites. And what 
that’s been taken to mean is, that this section of Genesis was a segment taken 
from some document which had not been written by Jews. And the 
importance of that is, that it’s the fundamental piece of information which 
provides for the possible historicity of Abraham. That is, if there was a 
document which was not a part of Israel’s traditions, not written by them, in 
which he occurred as a figure, then that’s as strong evidence as they’ve gotten 
that it was in fact some person, and not just the name of a tribe, or something 
else. (The term ‘Hebrew,’ itself, has other significances; it’s a very generic 
term in the ancient Near East, referring to a class of persons.) 

The thing is that here’s a category which has become crucial in biblical 
criticism because what’s seen about it is that a member would not use it to 
refer to himself, except under very special circumstances. It’s a category that 
we might say, roughly, is the possession of some group other than its users. 
In terms of current large-scale focus, the Muslims have a similar attitude 
toward the category ‘Negro.’ And what they want to do, in part, is to legislate 
the status of ‘Negro’ into having the same characteristic as ‘Hebrew,’ so that 
if you knew somebody said “so-and-so the Negro,’’ you could say about that 
somebody, “He’s not one’’ - unless it was a conversation in which the user 
was identifying himself to an outsider. 

‘Company’ is obviously the same kind of category. It’s used to refer to 
some outsider, or it’s used by an outsider to refer to himself when he’s talking 
to some member and characterizing himself as an outsider to them. I said it 
was a category in alternation to some other category. It is also a category that 
is used, by and large, by those who are using the alternating category to 
categorize themselves. 

We can examine lots of categories to see who is it that owns them. And 
we can see that that kind of question can be a really important one. We’ll 
see that when we consider what kind of difference there is between a 
category like ‘hotrodders’ and one like ‘company,’ and between, let’s say, a 
category like ‘hotrodders’ and one like ‘teenagers.’ We’ll also then see some 


Part II 

of the things that people are trying to do when, like the Muslims, they 
attempt to propose that Negroes shouldn’t use ‘Negro’ as a category, and 
when, like teenagers, they construct and attempt to use a category like 

For them to say that this fellow Bob is ‘company’ will increasingly be seen 
as an attack on, not simply the therapist, but in some ways, the ways that the 
institutions that get all of these guys here are constructed. Because in the first 
instance this category is one that the kids administer. It’s a generic category, 
available to whoever can bring it off, and that’s different than the category 
‘hotrodders’ which is laying around in this conversation. Let’s consider that a 
little bit. And in doing so, we’ll be moving more explicitly to a consideration 
of some of the things involved in the phenomenon of ‘an automobile 
discussion;’ what kind of topic that is, and what it is that can be accomplished 
within it. 

Now if we can take it that to some extent ‘hotrodders’ is a category that 
is by and large employed by kids to characterize themselves, and whose use, 
to some considerable extent, they enforce, and whose properties they enforce, 
and obviously it’s, at least to some extent, a category that rebellious persons 
can use, then at least one of the initial questions we might ask is: Why should 
it be the case that at least some people who go about doing kinds of rebellion, 
do it by formulating themselves as some particular type? That is, why do they 
set up a type? Why don’t they try to make themselves observable as 
‘individuals,’ so to speak? That might be an alternative to setting up a type. 
(Of course ‘hotrodder’ is by no means the only teenager type that they set up 
and enforce.) 

That kind of a question - Why do they set up a type? - is not necessarily 
a good question. What we want to see, in formulating the problem they have, 
is what is it that they seem to see as the things they have to come to terms 
with. And that seems to be in the first instance, that at least for certain 
activities which are rather important to them - like driving - it isn’t the case 
that there’s any big problem about what category will be used to classify them 
by whomsoever happens to see them driving around. Now, that is, in 
principle, a problem, and I've occasionally focused on more-or-less unique 
solutions to it. 1 And for kids in cars, they apparently take it that there is - and 
there’s probably good reason for that - a rather unique solution. And that is, 
a kid in a car is ‘a kid in a car.’ If a kid is driving, he’s seen as a teenager 
driving, and he’s seen via the category ‘teenager,’ compared to the variety of 
things that he could be categorized as. His problem, then, initially, is that he 
is in fact going to be typed; where, for one, the category ‘teenager’ is a category 
owned by adults. Which is to say that what it is that is known about the 
incumbents of that category is something that adults take it they know, and 
know without respect to kids’ proposing that it’s so, or agreeing that it’s so. 

The problem has some other aspects, which are perhaps more obviously 
deep. And that is, it’s clear that one of the crucial things to the teenager is the 

1 See Spring 1966, lecture 7, pp. 325-7. 

Lecture 7 


problem of independence, and one of the core facts about it is that in its given 
form, adults set up and decide and enforce what it is that it takes for a kid to 
be independent. And that poses for each kid a task that he has to solve on his 
own, in confronting an institutionalized setup that’s going to handle him one 
by one. Now the question is, is there some way of setting up solutions to the 
problem of independence, which kids administrate? So that they decide when 
you’re independent? Where adults will, in the course of time, simply be forced 
to accept those characterizations, as they’ll be forced to use whatever 
categories the kids invent to characterize themselves? I’ll consider the gains of 
that kind of thing, to some extent. But it clearly would make a considerable 
shift in the independence problem. For one, it would not be an issue where 
each kid faces the adult world on his own. To set up a type, then, like 
Hotrodder, or Surfer, or Beatnik, if it can be successfully done, is then to get 
for the collection of persons a very large gain. That is, if the ‘they’ group, 
whoever it be - call them ‘adults’ - come to use the type as well, but use it 
under the extremely important constraint that what it takes to be a member, 
and what it is that’s known about members, is something that the members 

Now a first thing that one wants to be able to do, is to so construct 
appearances, and to so let out information, that members can take it that 
when they’re seen by whomsoever, they will be seen as a member of the 
category they want to be seen as a member of. That is, they’re not going to 
be seen as ‘teenagers in cars;’ they’re going to be seen as ‘hotrodders’ - or 
‘beatnicks,’ or ‘surfers,’ or whatever else. That, of course, poses a series of 
tasks. What is it that it takes to get a category like ‘hotrodders’ across? 

There are some things that turn out to be quite obvious solutions to that 
problem. Everybody drives cars, and ‘teenagers driving cars’ involves nothing 
more than looking into the car, whatever car it is, and seeing that there’s a 
teenager there. Now the thing about hotrodders’ cars is that you don’t have 
to look into the car. The car itself is so constructed that at any distance you 
might choose, you see ‘a hotrod.’ And of course what’s crucial to that, in the 
first instance, is that other people do not play around with the products they 
get. They take the car and they drive it. 

The modification characteristics have some rather usual features for such 
kinds of categories, and that is, anybody can tell it’s a hotrod, but it is 
members who can tell if it’s a good one or a bad one, what rank it has, etc. 
And in that way of course, there’s strong protection against outsiders coming 
to see how it is that whatever ranking goes on, goes on — except by, in fact, 
becoming members of a sort. That thereby gives much freer play to the 
members’ deciding and enforcing what counts as a really good hotrod. Now 
there are, of course, rather classical reversal procedures, i.e., what looks like 
the worst car to anybody else, is the best car. And those kinds of things are 
very frequent in any such kind of operation. Presumably it’s just focussing on 
that thing alone, that ‘turn the world upside down’ in that sense, that makes 
almost all of, let’s say, the Beatnick characteristics. Hotrodders do not stand 
in that sort of simple opposition; it’s not a pure Hegelian operation. 


Part II 

Now if it’s the case that what goes into deciding the rank of a hotrod is 
something that kids decide, then one can at least begin to see how it is that 
they’re able to set up, and what are the kinds of things they would use to set 
up, machinery for social control over candidate members. That is, one is a 
member by recognition of others who are members. And thereby, to 
successfully get membership, you have to do what it is that they provide is the 
way to become a member. You don’t ask your parents for permission, and 
then treat that as your entry card. Nor is it something you can do without a 
very considerable commitment. It is not something that you can do on an 
occasion, in any given apparatus. That is, you can’t take your parents’ car out 
for a weekend and go hotrodding. There are a lot of things you can do with 
it, but that’s not one of them. And adults are, by and large, excluded from 
the use of these things. 

That has some rather important kinds of aspects to it. It permits one of the 
crucial things the kids are concerned to get, and that is, an initial equalization, 
such that the determination of one’s status as a member is not decided by as 
many relevances given by the socio-economic structure as can be excluded. It 
really is not any advantage to have, let’s say, the fastest stock car to start out 
with - and in any case, it would not be ‘hotrodding’ to drive it. And it’s very 
important, of course, that kids do not allow it to be ‘hotrodding’, and that the 
way to not allow it is simply not to recognize the moves of someone who, for 
example, pulls up to you in such a car. That kind of thing is very regularly 
done for all sorts of other things. Years ago, when sports car drivers used to 
wave at each other, there were all kinds of rules about what kind of car is a 
sports car; like a Yolkswagon would not be waved at. Volkswagons could go 
down the road waving at everybody, and get no response. Now what has to 
be seen is that the response is not anything that is enforced in some way, like 
if you wave at a Volkswagon somebody comes up to you afterwards, or 
anything like that. The whole group has to be defended, each one by himself. 
And in that way, Joe’s hypothetical little tale of the other fellow in a Pontiac 
station wagon is a kind of mutual delusion - if there is such another fellow; 
that is, they may be able to go through it together, but there is no proper 
member who would accept either of them. What Joe is proposing, I’ll 
consider eventually because for them it’s an extremely important kind of 
claim he’s trying to lay out. 

The character, then, of an attempt to set up a category like this is - let’s 
say within Western tradition - a classical attempt at how it is one goes about 
doing rebellion, at least the first feature of which is that one sets up a category 
you administer yourself, which others come to use, and come to use in just the 
unique fashion that they used whatever category they used on you before. 
Where, initially they used another category on you, uniquely. They didn’t say, 
“He’s middle class,’’ and “he’s this and that,’’ and “he’s such and such;” they 
said “He’s Negro.” And the same holds for the teenager. Now of course, that 
leaves those who don’t join, to be so called. 

Lecture 8 

Invitations ; Inexhaustable topics; 
Category-bound activities 

Given that we have gotten into a position where some aspects of how this 
group is organized, and how that organization is demonstrated, how they use 
it, etc., have been considered, we can return to the first part of this segment 
and dig some more out of it, which we’re now in a position to consider, and 
which can lead us further, also. 

Joe : We were in an automobile discussion 

Henry : discussing the psychological motives for 

Mel : drag racing on the streets. 

In our previous discussion of this segment we examined its collaborative 
aspects. There’s much more there than that, and I’ll deal with some of those 
things now. 

We said earlier that that first line could have been a sentence if it was just 
left as such. And that in not leaving it, Henry moved to use the structure of 
the sentence to give some indication that these people were organized. That’s 
grossly true, but there are other things about it. For one, if we consider the 
first sentence as a sentence, then what we want to ask right off is, what is being 
done by it? Among the things that could be done by it, one obvious thing is 
the following. Someone has just come in to this group, they’ve exchanged 
introductions, now there’s this first line. And it’s a very familiar first line, and 
what it is, is, a possible invitation. A bunch of people are sitting around and 
talking, somebody comes in, and they tell him what they’re talking about; 
where they can formulate that in such a way as to provide that it’s something 
that he could be talking about. It’s a topic for such as we, without any special 
classification of who ‘you’ and ‘we’ are, just by virtue of the fact that we’re all 
teenagers. And with regard to a consideration of the personnel, you’ll find in 
the rest of the conversation that there are several other places where Joe makes 
a move like that. For example, after they’ve worked Bob over, Joe introduces 
a question that is an invitation: “What school did you go to?,’’ and there may 
be others like that. 

What I take it they do is shut that invitation off by proposing, in part, that 
we’re talking about that thing which you know about, as any teenager would 
- but we’re not necessarily talking about it in a way that you’re informed on. 
But then there’s the addition to it, which is again kind of mundane and 
ordinary. You can make up sentences like this. Here’s one - it’s not exactly 



Part II 

in point, though I think it’s pretty close. Suppose there’s a bunch of girls - 
and we don’t need to specify the situation - and somebody comes in. One of 
them says “We were in a discussion about sex.’’ Then we get “discussing the 
physiological side effects of,” “using the pill,” as a third. The first and third 
are something that somebody might well figure that they could, just knowing 
that, come into. The second may, however, be something which says, this is 
something special and unless you have the special information that you don’t 
necessarily have, wait before you jump in. It’s the modification of the topic 
that undercuts the status of the preceding part as ‘an invitation.’ 

Given the kinds of remarks we so far made - and more I will make - as 
to the fact that the group is organized, we can get some pretty good idea 
about, on the one hand, an invitation should be made right off, and on the 
other, why it should be cut off. If the group is organized in a two-to-one setup 
(or even one-one-one, as long as it’s not three-to-zero), then it would be quite 
crucial to the members who this new candidate joins up with, and it could, 
of course, be quite crucial to him who he joins up with. And if he simply 
jumps into the conversation - where it’s how the conversation goes that 
provides those affiliations - then some of the kinds of resources that would 
permit the sort of decision some of them want him to make, would not be 

He doesn’t yet know that whatever he says is going to be treated as 
something which constitutes an affiliation to one or another side. And if it’s 
the case that even before this actual segment had got going they knew what 
the structure was, and it was two-to-one against Joe, then Joe’s best chance is 
to get this guy in, sight unseen. He might go with Joe. If so, things are 
perhaps equal, and if he doesn’t, maybe there’s nothing much lost, in that 
once the group structure becomes available it’s probably not unlikely that, if 
he makes his decision on that basis, he’s going to go to the two others. And 
that’s not unlikely, I’d say, not on just the sheer fact that any given person, 
if there were a two-to-one situation, would go with the two, but one of the 
obvious things about him is that he’s scared. He’s scared, and that’s why he’s 
here in part, and also, he’s scared now. And under that circumstance it’s likely 
that he’ll pick the safest way out; which he does eventually do. That is, the 
first remark which can readily enough be seen as taking a position on his own, 
is between 3:00 and 3:30, when Joe does one of his not infrequent coughs. 
Bob comes on with “Sounds like an old man.” That’s about as dramatic a 
remark as he can make in this. So in the collaborative sentence, then, it’s not 
the case that it really wouldn’t matter who it is that did any of the parts. It 
does matter, and the part that each of them chooses to use is not terribly 

There are some things about this segment that are worthy of at least further 
thought. But let me first make a remark about the segment that follows. The 
most extraordinarily interesting thing about it is the way in which the 
questions are informative. They say to him, “You must be crazy or you 
wouldn’t be here.” It’s a terribly important beginning because it says ‘We are 
that.’ Every question is a piece of information about the character of them; 

Lecture 8 


what it takes to be somebody who’s here - at least formulated in terms of 
what an outsider might say about them. 

And it’s very important that this question-answer series comes out right. 
If it doesn’t come out right, it’s not simply, as an obvious consequence, that 
he doesn’t belong there, but it may well equally be that they don’t. Because 
if he’s there, however it was decided, he could belong there. And if he’s not 
crazy, then, maybe this is not a place for such, and if it’s true about them, then 
they don’t belong; where there’s a notion that the people here ought to have 
problems they have in common. (I’m not giving any normative formulation 
to ‘problems.’ They’re called ‘problems,’ and they would say they are called 
‘problems’ without necessarily buying into that on their own part.) But if it’s 
the case that they go through this procedure of asking a set of questions, and 
they’re right, then they get a very gloomy finding, and that is that the society 
is right, because these are a list of complaints that are made about them. If, in 
fact, they locate whomsoever happens to come in here, then there’s some sense 
to the fact that they know what they’re doing in putting people in such a place. 
So the procedure is an extremely important one, however it may come out. 

Let’s go back to this first segment. They’re in an ‘automobile discussion.’ 
Now here’s a fellow coming into this therapy situation, very likely his first. It 
could be supposed that he has no idea what it is they would do in there. If 
they wanted to put him on, they could name any topic they pleased, and he 
probably would have to figure that that’s what they might be doing. When 
‘automobile discussion’ comes out, it could be a very nice thing to hear. “Oh, 
they talk about whatever it is kids talk about anywhere, and they’re not going 
to be talking about things that are really godawful, or for which I just don’t 
have any information so that I could partake.” Except it’s much worse than 

We get the modification which says hold on for a while, see what we’re 
talking about and how we do it. And then we get a sample. And the sample 
can be seen partly as a kind of ostensive definition of what ‘discussing the 
psychological motives for drag racing’ means, and how a topic that you can 
name in any scene that kids happen to be talking in, gets handled in this 
scene. In which, we could say grossly, that what might appear to be as safe 
a topic as you could pick, turns out to be altogether unsafe. And whatever 
position you take is nonetheless usable to find out what’s bugging you. And 
I suppose as strong a demonstration as needed to show that once you’re here 
there’s no place to hide, is to pick as mundane a topic as you could name, and 
then use that to carry on the intimacies of examination that they do. 

When I use the term ‘safe’ for ‘automobile discussion,’ I’m not simply 
saying that. That’s how they would quite consciously formulate the matter. 
At one point in another session they’re for some reason or another exchanging 
fantasies about something they’d like to have. It’s about maybe Bob’s second 
session there, and he says he’d like to have a “’64 Triumph Bonneville,” and 
he gets a lecture from Henry as to the fact that that’s not the sort of thing to 
pick; that you don’t have to pick anything that sounds as ordinary as that, and 
if you think that by picking something that sounds safe you get away with 


Part II 

anything, you won’t. (Which is not to say that as far as any of them might 
well understand it, fantasy is about such things.) 

I just want to extend the matter a little further, and simply suggest a kind 
of problem about which I really don’t have very much to say, but which is, 
I think, examinable. Consider things like the topic ‘automobile.’ One of the 
things that an outsider, an adult, somebody who is not American, expectably 
says is, well that could be a topic, I suppose, but it would probably be readily 
exhaustable; how much is there to say about it? And you could imagine that 
they could construct such a notion of cars as a topic by considering, for 
example, that there are a finite set of parts to cars as facts, and that, as you 
might exhaust the parts if you were assembling it or taking it apart, then you 
could exhaust the topic. That is to say, you could have an image of eating into 
a topic as you discussed it. 

Now there are lots of topics for any sets of persons, but within cultures 
there are topics that are intrinsically rich, in the sense that whatever it is that 
members of that culture tend to talk about - that is, whatever themes they 
talk about - they can talk about via that thing. For example, an automobile. 
If one sat down to make a list of what things members of this society can 
talk about when they’re talking about something, then one might find that 
for kids, automobiles are like that. For example, ‘independence.’ You could 
make a list of themes like that, and then see whether or not a conversation is 
had, where that’s the theme that gets focussed about automobiles. If this is so, 
then you have a rather decent way of saying that there is to some extent an 
independent culture operating. And that the automobile has now become a 
focus for kids in the way that whatever else is a focus for the adult culture - 
if ‘adult culture’ is a thing to talk about. 

Of course it might not be the case that some other culture has a single 
focus - if cars are a ‘single focus.’ What I’m trying to say is that one of the 
things one has to do in examining this phenomenon of automobiles as a topic, 
is to formulate the character of that object within the set of things which can 
be discussed, for whatever you’re going to formulate it that people are doing. 
And then of course it would be quite irrelevant to see that adults can’t or 
don’t or won’t say such things about cars. Cars are now an object similar to 
parrots for a society in which a parrot is a god. And you wouldn’t say, “Well, 
what can you say about a parrot?’’ or “What kind of interest could people 
have in the difference between this parrot and that parrot, or this history of 
parrots, or whatever else?” You would just not know what they’re talking 

You get, then, a way of seeing the kind of split that would take place when 
an object is made into a sacred object - and a quite mundane object can be 
made into a sacred object. Where, for others, it’s a purely secular thing: You 
take it as you get it, and you drive it. And whatever happens to it, so what? 
You can always get another. And also, of course, you get a way of 
understanding that persons who deal with this as perhaps sacred have feelings 
about the miserable way anybody else deals with it. But I don’t really know 
that automobiles are a first-order object. It could be, because most of the 

Lecture 8 


conversations I listened to, I found so terribly boring that it’s only recently 
that I’ve gotten up enough strength to sit and listen to them again, when they 
go on for six hours at a shot, talking about carburators. 

There’s a classic story about Malinowsky going into the field, wanting to 
find out about geneology. He would talk to the natives about geneology, and 
he found out that each native would talk for an hour before giving him any 
geneology, in just bragging about them, and he’d have to go through a whole 
big thing. Then at one point he remarks that he really became an 
anthropologist when he could realize that geneologies were irrelevant, and it 
was the talk, the bragging, etc., which was crucial. And you have to learn to 
see that happening. I used to sit and listen through these automobile 
discussions for the points when they would overtly talk about their troubles, 
and I just don’t think that was at all right. It may be its partial distance from 
somebody like myself that makes this a really good means for examining 
what a topic like that is, and what you would be talking about when you say 
it’s inexhaustably rich. 

Now I want to move to considering the argument that Joe develops, 
because one of the things we come to see, then, is what kind of category is that 
category ‘poor little rich kid’ which is thrown at him when he develops his 
position. But I can’t do that in the time we have left today, so what I’m going 
to do is something that I could have either taken the whole course to do, or 
I’ll have to do in ten minutes; it’s either that complicated or that simple. So 
I might as well do it this way. 

One of the clear problems that I can propose from this conversation is, we 
could ask: How does it happen that discussions which become arguments end 
up in name-calling? That’s a very standard observation, that it happens. It 
happens here. Now, how does it happen? There are several things involved in 
the formulation. We’re assuming that the discussion becomes an argument, 
and we’re not asking how does the discussion become an argument. We’re 
saying that once it becomes an argument it can end up in name-calling, and 
that ends it. 

The way we go about handling it involves some very general kinds of 
observations which only become really interesting when we work them in 
detail - which we won’t get to do. 

One of the ways that a problematic occurrence is resolved, is by assigning 
to the doer of it, some category about which it can be said that the activity 
done, is ‘bound’ to that category, i.e., if you knew in the first place that he was 
a such-and-such, it wouldn’t be any problem as to why he did the thing he 
did. ‘They’ do such things. Now that fact has consequences in all directions. 
One of them is, for example, if a problematic occurrence has happened, and 
one knows a category that’s bound to it, i.e., where that category would be 
said to be the person to do the thing, then you can construct a search 
procedure for finding who, in fact, did it: Look to the set of people who are 
so categorized. You could also, apparently, determine that any person who is 
proposed to have done it, did it or didn’t. If he isn’t a member of that 
category, then he wouldn’t have done it. And that has all sorts of uses, and 


Part II 

there can be lots of shifts in it. So, for example, the basic character of the 
argument about Oswald’s murder of Kennedy turns on the issue of some 
people saying, “Of course if he’s a Communist he would have done it,” and 
others saying, “The last person who could have done it would be such a one 
and therefore he couldn’t have done it.” To which, as I mentioned earlier, his 
mother had the best solution of all, when she said that he’s a human so he 
could have done it - which is to say that it’s not something that is to be found 
problematic in the way everybody’s been posing it. 

The fact that some activities are bound to some categories is used, then, 
in a tremendous variety of ways, and if somebody knows an activity has 
been done, and there is a category to which it is bound, they can damn well 
propose that it’s been done by such a one who is a member of that 

What’s important, in part, is that it’s not the case that deviant activities are 
especially problematic, but there are categories of persons who do deviant 
activities and you’ve got a solution to a deviant activity if you’ve got a 
member of a category which is known to do this. I’m not going to deal with 
the problem of exceptions, although it’s dealable with and there are some very 
nice things about it; I’m just not in a position right now to lay it out. It is the 
case, though, that exceptions just don’t matter. That’s easiest to see by seeing 
that the first Negro and the first ten Negroes you know, can be seen by you 
to be exceptions to what you know about Negroes. It’s not the case that 
exceptions involve any change in what you know about the category’s 
members. For all the categories that have such kinds of characteristics as that 
there are a bunch of activities bound to them, exceptions don’t matter. It’s 
built in that there are exceptions, and they just don’t affect what you know. 
You know that category does the following, and you know that there are 
exceptions, and they do not involve you in modifying what you know. I talk 
about that as: All these categories and the things that are known about them 
are ‘protected against induction.’ 

The question is, what makes a problematic activity? And an utterance, of 
course, can be a problematic activity. One of the simplest ways that an 
utterance can be a problematic activity is, if a set of them has been made by 
a set of persons and they’re not consistent. That is, there’s a position 1, a 
position 2, etc. Now if, let’s say, you have two positions that are inconsistent 
to each other, then it may be the case that you can make one position stand 
by simply removing the other; like providing it’s problematic and explaining 
it. The other then stands as the only one left. 

Now then, one of the ways that one goes about making a problematic 
position accounted for is, for instance, to assign some category to its sayer. The 
category is, of course, important, because there can be some categories that 
are, for example, the categories of persons who are entitled to say such a thing. 
So that if you say that some scientific assertion is made by a scientist, that 
would not explain it away, but will provide that it’s correct. But for a great 
many, if you assign a certain sort of category to it, then you can provide that 
the thing itself can be explained away. And that simple operation has been the 

Lecture 8 


basis for one quite sophisticated sociology. Mannheim proposes that any 
sociological assertion can be given that formulation. That is, there are a set of 
categories in the social structure, all of which are competing in that social 
structure: Upper class, lower class, bourgoisie, worker, whatever you want. 
Take any sociologist’s assertion, and say he says it from one or another of those 
categories, i.e., assign a sociologist to a category, and you’ve made his 
statement subjective. 

Lecture 9 

Character appears on cue ; Good 
grounds for an action 

I’m going to start considering : 1 

Joe\ In that Bonneville of mine. I could take that thing out and if I’ve got 
a tie and a sweater on and I look clean? ninety nine percent of the 
time a guy could pull up to me in the same car, same color, same 
year the whole bit, roll up his pipes and he’s in a dirty grubby tee 
shirt, and the guy’ll pick the guy up in the dirty grubby tee shirt 
before I - he’ll pick me up. 

This is an extraordinarily complicated statement and we’ll attack it over a 
while, I suppose. Let me just begin by noticing a relatively minor thing. 
When we get this thing beginning with “In that Bonneville of mine’’ we have 
what is a relatively rare, but standarized occasion, and that is that we can say 
right off that ‘Bonneville’ is going to be the N of some N — *■ P sequence. It 
looks for sure that that thing is not going to be dropped, and that we’ll then 
get ‘it’ or something like ‘it’. In our discovery of the rule, I said normally we 
do not start with the N, we start with the P. But in this sort of case, we can 
pretty well be sure that we’re going to have the reverse rule apply. That, of 
course, involves the fact that we get this kind of prepositional phrase which 
would follow , e.g., “I could go out ... in that Bonneville of mine.’’ 

Now, generally throughout this statement we see what you might imagine 
to be a kind of queer way that the whole thing gets assembled, i.e., clause by 
clause. There’s no sentence in it anywhere - though one might consider from 
“ninety nine percent of the time ...” on, to be an adequate sentence - but 
otherwise you get a buildup (I’m not using that in the same sense I used 
‘buildup’ as a technical term previously) in which, at the end of any given 
clause it’s nonetheless the case that we still haven’t got a sentence, and we 
clearly still haven’t got the end of an utterance. To what extent he uses clauses 
(and there are much more elaborate things he does) to preserve the fact that 
you can’t stop him at any point and have heard what he’s going to say, that 
may be one of the things he’s engaged in doing here. 

(If you’re interested, I can mark it for the way he said it: “In that 
Bonneville of mine.” There’s a pause. “I could take that thing out” pause 

1 The fragment is transcribed from the lecture and is not exactly the same as the 
version in the handout. 


Lecture 9 


“and if I got a tie and sweater on and I took clean’’ pause “ninety nine percent 
of the time a guy could pull up to me in the same car” pause “same color 
same year the whole bit” pause "roll up his pipes” pause “and he’s in a dirty 
grubby T-shirt” pause. And then the rest of it runs out to the end. Then he 
starts up again, but gets cut off.) 

One of the things one has to examine about that statement is, how is it that 
it’s nonetheless a quite understandable utterance, i.e., despite its grammatical 
oddness and despite (if you want to use the word ‘despite’) some other 
features it has. For example, the phrase “ninety nine percent of the time” 
occurs in juxtaposition with “a guy could . . . etc.,” except I don’t take it we 
hear it as modifying “a guy could . . .” If you were to put it in sequence, it 
would be: “ninety nine percent of the time . . . the guy’ll pick the guy up,” 
and we nonetheless take that clause and put it where it belongs, when we 
come to understand it. So there is, then, a considerable freedom in the 
positioning of the parts of this statement, where that freedom apparently 
doesn’t undercut seeing what it is he’s saying. 

It’s also an enormously condensed statement. And that’s one of the things 
I want to begin to focus on. Let’s take as a given, for the moment, the fact 
that in “the guy’ll pick the guy up,” the first reference doesn’t refer back to 
the prior “a guy could pull up to me . . . ,” though that’s a perfectly 
expectable thing, but that we see perfectly clearly that it doesn’t, but that it’s 
a new fellow, and that it’s a cop. 

Now, if you remember back to the beginning of the course when I 
considered “The baby cried. The mommy picked it up,” one of the things I 
was remarking there was that when a character who has some proper grounds 
for occurring and some proper thing to do, has its cue, then there’s no need 
to account for how they happened to have come on the scene. “The baby 
cried;” you don’t have to get a characterization of where the mommy was, 
that she was anywhere around; the mommy does her job. 2 You can notice 
again here that the cop is not introduced, as, for example, is this other guy 
who is driving - he pulls up, gets described, then we have this bit. What 
occurs is good grounds for the cop to do what he ought to do; he’s on the 
scene and he does it. So he’s introduced via the action he does, where the 
grounds for that action are laid out, though how he happens to be there need 
not be indicated. 

And the question of what character doing what action that kind of thing 
can hold for, is quite non-trivial. To provide for this introduction of the cop 
(and, for example, that introduction of the mommy) being quite unprob- 
lematic, we would have to work out analyses which would involve building 
the characteristics of a class of actor - in this case, ‘cop’ - as a staff deputed 
to do a set of jobs, and describing how they go about doing those jobs - for 
example, by just driving the highways - and then considering the information 
about such a class of actors that any member would have, such that it would 

2 This particular consideration was not captured in Sacks’ notes for lectures 1 and 
2. For a similar consideration, see Spring 1966, lecture 2, p. 254. 


Part II 

in no case be problematic that on a single occasion of the cue occurring, the 
character occurs. (This is not to say that the sheer fact that there’s a norm that 
says speeders ought to be arrested is sufficient to provide that the fact that he’s 
speeding brings the cop on the scene.) For organized crime it would be quite 
different. A bank robber would be much puzzled how come when he arrived, 
the cops were there waiting for him, and he would ask, “How did you know?” 

One can presumably consider this sort of phenomenon and perhaps get an 
idea of what the sense was for, let’s say the Greeks, who introduced in the 
tragedy, the appearance of a god at some point. We might come to see it, not 
as what we give it the sense of, i.e., as absolutely unwarranted or unexplicable, 
or just happening, but as something that was somehow institutionally 
provided for. We would want, of course, also to see, then, what sorts of 
proper occurrences need an account of how a character who is doing what’s 
expected or proper, comes to be there. We could perhaps dig out the different 
classes by just examining things like this statement of Joe’s, where it’s 
absolutely unproblematic that he doesn’t introduce the cop as sitting behind 
a billboard, or whatever else. The cop is introduced via his action. 

And in this case at hand, it’s a rather more powerful kind of thing than 
even that. In “The baby cried. The mommy picked it up,” we had the 
specification of what was in fact good grounds for the mommy to do her 
action; “the baby cried.” In this case, that, too, is just left out. It is not the 
case that rolling up his pipes is what’s good grounds for the cop to arrive. It’s 
that rolling up his pipes is the first move in a sequence which the occurrence 
of that technical term provides as going to happen. The driver pulls up, 
signals the other, they have a kind of drag race, and the cop catches them in 
the middle of it. That part is dropped out altogether here; filled in by the 

Now it’s very nice to have the hearer fill it in because one of things that that 
will guarantee is - and this is an important part of this fellow Joe’s claim - 
if you’re going to fill it in, you’re going to fill it in with what would be 
legitimate grounds for that arrest to take place, and not with some quirk or 
some erroneous arrest. Then the question of what can be left out of an account 
- especially where it’s the thing that’s left out that provides for the crucial 
action to happen - is another rather crucial thing to consider. For one, that’s 
the way that the condensation of an account can take place. The work of 
digging out what’s being said is, then, the task of the hearer, and one can 
leave out what one knows the hearer has to fill in in a specific way. That is, 
the hearer is not going to say “Rolling up his pipes is nothing to get arrested 
for,” but there is some sequence, of which that is the first move that occurs, 
and when the actionable move occurs, then the arrest takes place. 

In fact, it’s not simply the case that a hearer can fill in such a blank. They 
can do that, but they can do a lot more. Here is a piece of conversation which 
has the following kind of setup. A fellow is calling a counselling agency in 
regard to the trouble he’s having with his wife. It’s quite early in the 
conversation; a first conversation between two people who don’t know each 
other. ‘A’ is the staff member, ‘B’ is the caller. 

Lecture 9 


A: Yeah, then what happened? 

B: Okay, in the meantime she says, “Don’t ask the child nothing.” Well, 

she stepped between me and the child, and I got up to walk out the 
door. When she stepped between me and the child, I went to move 
her out of the way. And then about that time her sister had called the 
police. I don’t know how she - what she — 

A: Didn’t you smack her one? 

B: No. 

A: You’re not telling me the story, Mr B. 

B: Well, you see, when you say smack you mean hit. 

A: Yeah, you shoved her. Is that it? 

B: Yeah, I shoved her. 

Again, it’s not like they’re old buddies. The secondary characters here are just 
named categories, i.e., ‘her sister.’ He doesn’t know her sister, doesn’t know 
if she’s short-tempered or what, doesn’t know how B and his wife get along, 
or what B is wont to do if his wife does something. Nonetheless it’s not simply 
the case that he has some idea of what it is that would provide for the sister 
to call for the police, but when he’s told it’s not so, he takes it that there’s a 
lie going on, and he’s in a position to insist and not simply to guess at what 
it is that transpired. In that sense, then, knowing what are good grounds for 
the police to come is something that permits him not only to fill in a blank, 
but also to deny a statement which proposes something else, with no need to 
specify anything more than ‘me and my wife,’ ‘her sister,’ ‘the police.’ What 
is in operation here is the pure categories, and their organization, and what’s 
known, given their use. The guy who calls is in no position, apparently, to say 
“How do you know?,” “I’m telling the story,” or whatever else. It’s not 
simply then that the guesser feels quite assured in proposing what it is that 
happened, but the other guy is just as well in a position to see that the guesser 
had good rights to do that. I take it, then, that one of the things we would 
not be saying is, it’s a matter of style whether one leaves out a sentence that 
might be put in. 

How that can happen is of tremendous importance, and it tells us a great 
deal about what it is that, among a set of persons who don’t know each other, 
nonetheless can be taken as so about how the world operates, in its 
particulars. That is, on a single occasion. It’s a different kind of business than, 
“If you go hotrodding you’re bound to get caught.” It’s on this occasion the 
cops arrive. How come? 

Now, with such a guide, you could begin to see such issues as, what the 
task would be of a foreigner to understand that thing. Suppose you were to 
translate this. The issue would not be anything like, do you come up with a 
word-for-word translation? What you have is do is build in characterizations 
of the way a set of institutions operate, the way sets of persons know what 
anybody else knows, and the like. Just to provide for a quite ordinary segment 
of conversation. 

The big question is how we come up with clean ways of formulating what 


Part II 

the kinds of condensation are. You could ask it in a fashionable way: What 
would it take to have, as the staff member in that last conversation, a 
computer? What would it have to know, to be able to engage in a little 
conversation like that? If you could build computers that could gossip, then 
you would have a formal characterization of what a culture was. And the 
character of what these condensations are, is not to be supposed as satisfying 
any unexamined notions of what, let’s say, we take to be prototypical kinds 
of condensed talk, like talk in mathematics. That’s in part the problem with 
Reichenbach’s seventh chapter, where he simply will suppose that when 
persons use terms that have a known logical sense - like ‘and’ - then that’s 
the sense they have. 

Returning to “the guy’ll pick the guy up.” Earlier I mentioned this notion 
that some activities are bound to some category. And I take it that when we 
hear “the guy’ll pick the guy up” we use ‘pick up’ as an activity that ‘cops 
do, to provide that ‘the guy’ is ‘a cop.’ And we don’t find ourselves in the 
position of being confounded by the antecedent rule, which would say ‘the 
guy’ has to be the fellow introduced as ‘a guy,’ since ‘a’ followed by ‘the’ is 
a perfect way of indicating that the same person is being referred to. 

There’s a kind of elegance to the use of ‘the guy’ there. It is in the first 
instance a tremendously awkward kind of way to say the thing, “the guy’ll 
pick the guy up.” To have it that ‘the guy’ here is not the person who is 
introduced as ‘a guy’ earlier, would seem to be generative of further trouble. 
But once he’s got this ambiguity going, he has an opportunity to reintroduce 
one of the crucial assertions of the whole segment, and that is, “in the dirty 
grubby T-shirt.” So that grammatical awkwardness provides the occasion for 
emphasizing what he’s trying to emphasize, where, were he to do it in the 
grammatically proper way, such an occasion would not be available as it is 

And we can find that ‘dirty grubby T-shirt’ is something that is 
emphasized, independent of its being repeated, in the following ways. First of 
all, the whole statement is hypothetical. Every clause in it is hypothetical. In 
the whole statement, there is one use of the most standardized way of making 
a hypothetical statement, that is, the ‘if ’ in ‘. . . and if I got a tie and a sweater 
on ... ” What that single usage seems to involve is, given that there’s a 
whole set of things that are introduced as precedents to someone’s getting 
picked up, how is it that he’s going to focus on what he takes to be the crucial 
differentiating thing for that happening? It seems that he marks that crucial 
contrast with ‘if.’ Where, then, we can see as parallel construction, “and if I 
got” and the subsequent “and he’s in;” that is, we get not simply the 
difference between my being clean and his being dirty, but the parallel 
construction (‘and if,’ and then again ‘and if,’ with the ‘if’ dropped out - and 
you don’t have to have the ‘if’ in such a parallel construction) sets those 
things up as what differentiates. 

Furthermore, ‘dirty grubby’ is an emphatic construction in itself, in a 
variety of ways. First of all, it’s a double adjective; the same thing being said 
twice, which is kind of characteristic of kids, and characteristic of this fellow, 

Lecture 9 


also. Then, it’s not just a doubled adjective, but a doubled one that is 
soundwise emphatic. ‘Dirty grubby,’ ‘teeny weeny,’ things like that. 

There are other things involved in this, as well. ‘Grubby’ happens to be the 
way that one of the members of this group gets referred to. A way they get 
at Henry is by characterizing him that way, and apparently he does, indeed, 
tend to be grubby. On page 5 of the eight-minute segment, he says, just 
before 6:30, “I’m not grubby’’ and that brings a laugh. Now, I’ve been 
leaving that sort of thing out; that is, the way that in any statement which is 
not at all specifically referring to somebody in this scene, they are nonetheless 
doing that. That would have to involve providing much more transcript than 
the eight-minute segment. 

Lecture 10 

Clausal construction; Hotrodding 
as a test 

I’m going to continue considering this thing: 

Joe: In that Bonneville of mine. I could take that thing out and if I’ve got 

a tie and a a sweater on and I look clean? ninety nine percent of the 
time a guy could pull up to me in the same car, same color, same 
year the whole bit, roll up his pipes and he’s in a dirty grubby tee 
shirt, and the guy’ll pick the guy up in the dirty grubby tee shirt 
before I - he’ll pick me up. 

I’m not in a very good position to lay out how, in its details, this utterance 
gets constructed of a bunch of clauses and phrases rather than a set of 
sentences, or what each of the variations on what we might take to be the 
normal form for such a set of assertions is doing - if it’s doing something. I 
pointed out some things earlier, and I can point out a couple more. 

One of the rather obvious facts is that, given this way of constructing his 
utterance, the clauses are adding up and it’s not the case that at any given 
point - even at the point where we have that question mark (“. . . and I look 
clean?”) - that he’s said what he has to say. And that’s observable by the 
hearer. The hearer doesn’t have to guess that he has more; it’s observable that 
whatever it is he’s done, he’s not yet finished with it. There are some quite 
clear ways that the fact that the utterance is still open at any given point, is 
done by him. For one, whatever much we might have to say to give an 
adequate characterization of the semantics of ‘if,’ it does seem clear that there 
can be a multiple set of things being done in a statement that contain, at an 
early point in it, an ‘if’ clause. It clearly is a marked first occurrence of some 
sequence in which there will be a second part, the ‘then’ clause. There may be 
an overt ‘then’ clause, or a clause that turns out to be a ‘then’ clause. Where, 
then, as long as the ‘then’ clause hasn’t occurred, the utterance remains open, 
and one can wait to examine any given further clause to see whether that is 
the ‘then’ clause, and can close the thing off. 

Or, for example, we could ask what would be the difference between 
saying ‘‘I could take that Bonneville of mine out” as compared to what he 
does, ‘‘In that Bonneville of mine. I could take that thing out,” where it’s 
reasonable to say that neither would be heard as a complete utterance. Now, 
given that we can see, when ‘‘In that Bonneville of mine” occurs there’s going 
to be a second reference to the car, e.g., ‘it,’ or as in this case, ‘that thing,’ then 


Lecture 10 


perhaps what he is doing with this construction is having that second part 
provide a contrasted reference to the earlier mentioned car. That is, “If you 
take a big fancy car out’ ’ is what he’s introduced in his last statement. ‘That 
thing’ can then be, perhaps, not only a reference to the Bonneville, but a 
contrasted reference to the other - where a ‘big fancy car’ is a hotrod. That is 
something that could be done with a clause like that, as part of a pair, ‘this 
and that,’ ‘that thing’ ascomparedto ‘theother.’ and it would make part of his 
argument, which is, “If you go out hotrodding with a hotrod you’re bound to 
get caught” and in contrast, “If you take the Bonneville out, you won’t.” 1 

There are, further, a set of ways that this thing is constructed which are not 
exactly consistent with what is proposed to be the way that certain connectors, 
most particularly ‘and,’ get properly used. A consideration of those can be 
found in Fries’ book The Structure of English. His grammar is terribly 
important because I guess it’s the only grammar of English that was 
constructed by reference to an attempt to handle actual conversation itself. 
Fries sat down with a bunch of telephone conversations and built his grammar 
out of that. Quite unique in that way. 

One of the things he proposes about ‘and’ is that the term on either side 
are from the same form-class. That means, roughly, adjective ‘and’ adjective, 
verb ‘and’ verb, etc. Seymour Chapman then extends Fries’ form-class use of 
‘and’ to handle similary constructed clauses. That's in a paper called 
something like ‘English sentence connectors,’ in a book edited by Marquand, 
entitled Studies in Language and Literature. Now, what we have in this 
statement of Joe’s, are clausal uses, except that they’re kind of queer. Take a 
sequence like this: ‘. . . roll up his pipes and he’s in a dirty grubby T-shirt and 
the guy’ll pick the guy up . . .” The consecutive clauses do not seem to be 
similar, but they are separated, in a way. That is, there those pauses: “roll up 
his pipes” pause “and he’s in a dirty grubby T-shirt” pause “and the guy’ll 
pick the guy up . . . etc.” We could treat those pauses as putting “and he’s 
in a dirty grubby T-shirt” in parentheses, and then we can notice that “roll 
up his pipes” and “the guy’ll pick the guy up” are of the same class, and 
‘and’ connects them, i.e., “roll up his pipes . . . and the guy’ll pick the guy 
up.” The ‘and’ preceding “he’s in a dirty grubby T-shirt” is connected to 
“and if I got a tie and a sweater on and I look clean,” and there it does the 
same sort of work. So what we’ve got is non-consecutive uses of those ‘and’s, 
which nonetheless do the same work that they would be doing if they were 
consecutive - and how he comes about to break these things into this 
non-consecutive fashion, I can’t really say. But what we’ve got is a series of 
parallel clauses: “If I got a tie on” and “he’s in a dirty grubby T-shirt,” “I 
could take that thing out” and “he could pull up to me,” etc. We’ve seen 
earlier that this guy uses these parallel constructions very much, e.g., “you’re 

1 Some of the class disagree with the proposal that the distinction is between 
‘hotrod’ (‘big fancy car’) and ‘Bonneville.’ They take it that, as one of them proposes, 
although the “Bonneville really isn’t a ‘big fancy car’ (i.e., a ‘hotrod’), this kid is 
trying to pass it off as one.” To that, Sacks replies that if they had access to more of 
the transcript they could see that the distinction holds. 


Part II 

bound to get caught” and “you’re bound to get shafted,” and a whole series 
of others like that. What they are, I don’t have any idea. 

In Joe’s first statement, he said “If you take out a hotrod and you go 
hotrodding you’re bound to get caught.” Now what he does is to substitute 
for the hotrod this Bonneville station wagon, and he here reposes the 
circumstances of getting caught, where both persons are using “the same car,” 
and what differentiates them, and what turns out to be relevant to getting 
caught, is how they’re dressed. So we have a sequence of moves: It’s not 
simply, now, that they’re not driving the hotrod, but even driving the 
Bonneville there is a chance that they’ll get caught, and the way that gets 
reduced is by what he wears. 

What I want to do now is try to lay out, to some extent, what some of 
the conflicts are between the guys in this discussion, and what things they 
are proposing. I partially considered what the other guys are proposing a 
couple of days ago. What he’s arguing, as I see it, is that your basic aim 
ought to be to go speeding while minimizing your chance of getting 
caught. And to minimize your chance of getting caught, there are a set of 
constraints which are put on how you go about doing it. If you don’t use 
them, you can’t win; so he proposes. From the hotrodder’s point of view - 
if you can take the others to be espousing that point of view - such a position 
misconceives what they’re up to. What they want is not a guarantee that they 
won’t get caught. If they had that, they would have lost right then and there. 
They want a situation in which they can expect to lose, because what that 
permits is, if they happen to win, the formulation of explanations for having 

That is to say, if it’s the case that not many people get picked up in a 
Pontiac station wagon, then if you don’t get picked up in a Pontiac station 
wagon, there’s no problem which needs an explanation. If, alternatively, most 
people who go hotrodding get picked up, then on any given occasion when 
you don’t get picked up, there’s a problem which can be explained, and which 
can legitimately be posed as, “How did that happen?” At that point, and into 
that problem, can be fitted ‘skill,’ as an explanation. 

It’s in that regard that why it is that one chooses an activity like 
hotrodding, that involves violating some legal norm which is an important 
legal norm, can be fairly clearly seen. What you need in order to get the 
possibility of the proving of skill, is a situation where, if some given activity 
has been done, ‘A’, then either ‘B’ or ‘not-B’ are both observable. That is to 
say, both ‘getting caught’ and ‘not getting caught’ are observable things, so 
that you can say that ‘not getting caught’ is something that happened, or that 
‘getting caught’ is observably absent. 

And of course an enforced legal rule is a rather usable object for this test of 
skill. So it’s not at all the case that if they could be guaranteed that they would 
not get caught, that’s something they would want. It’s a misunderstanding to 
suppose that drag racers are simply playing each against the other. They’re 
play against the cops just as much. And they need the cops, because they need 
the cops’ catching or not catching of them to formulate that they have done 

Lecture 10 


the thing correctly; to even pose the fact that something has happened which 
they could have done correctly. In order to have a situation where you need 
an explanation, so as to be able to put in the explanation you’d like to put in, 
then what you need is some situation where these B or non-B alternatives are 
both equally observable things. If not many people get picked up, then if you 
don’t get picked up, it’s just that not many people get picked up. 

Now the confrontation with the police is not simply accounted for in this 
way, though that’s a good part of it. There are some other things, and if, 
when I’m talking about them, I use a kind of glorified language, let me just 
say that it’s the language the kids use. These kids, and other kids who are 
routinely engaged in observably deviant activities, regularly formulate the 
problems in the same way. They start out with the fact that, in just cruising 
the highways, they are observably deviants. That is to say, they are seen as 
‘hotrodders in their cars,’ and they are seen as persons who either have just 
finished, or are just about to, or in any case are routinely engaged in, speeding. 
The rules of the game are that anybody who speeds can be arrested, but 
they’re in a situation where, further, they know that the cops can, if they care 
to, arrest them at any time. That’s not simply because the cops can say that 
they were speeding, but because their cars, as with anybody else’s car pretty 
much, if stopped and examined by reference to the criminal code, will be 
found to have one or another technical violations. As they put it, “Last 
weekend they were arresting people on Hollywood Boulevard for dirty 
windshields.” And they know that anybody other than those who are 
observably deviant will not get arrested, unless there is such an occasion as 
speeding. What they can then examine is whether the police will conform to 
the rules that the police ought to conform to, i.e., treat them as they treat 
anbody else, or whether they’ll succumb to the temptation to violate the rules 
of the game and make an easy arrest. 

And in their observable deviance, then, they have a means of learning 
whether or not people like the police are worthy of respect - whether they 
conform. And if they don’t, then of course they can say, “They don’t 
conform, why should we conform?,” etc., etc. In any event, you’ll find that 
they’re terribly carefully oriented to the fact that these others do or do not play 
according to the rules of the game. And the best way they have of testing that 
is, of course, to put themselves in the situation where they give a deep 
temptation to the others, to violate. Then if they don’t, you can say they’re 
respectable. They get a view, then, of organizations like the police, which 
others don’t have available. And it’s a very persistent topic as to whether, and 
how, the police do or do not behave as they ought to, or whether they just 
pick you up because they know they can pick you up - since, for one, if 
anybody sees them pick you up, the others all have an explanation: You must 
have been hotrodding. (Negro gang kids have exactly this same orientation. 
They engage in long complex discussions, which are morality discussions of a 
standard sort, as to how the police behave, and who is worthy of respect.) So 
these kids have a way of checking out the extent to which that moral order 
that they’re supposed to conform to is followed by those who are its enforcers. 


Part 11 

They’ve got a way to test it, they test it with perfectly good will, if they get 
arrested for speeding, that’s the way it goes. 

There’s really, then, a tremendous ethical conflict going on between that 
position and what Joe is proposing - which is something like an adult 
speeder’s view; that is, if you look perfectly square then you’re in a position 
to get away with violations that others would never get away with. And he’s 
proposing, then, to use that cover of his clean appearance, to generate 
violations while being protected from being caught. For him, then, if he were 
to do it, he would be gaining lessons in cynicism. It is also the case that what 
he’s proposing is not available to any given kid; that is, he’s proposing to use 
a set of resources that any given kid would not necessarily have. And their 
concern is to try to set things up so that any given kid would have a chance 
to participate in just the way that any other would, and, of course, to get a 
view of how that legal order is operating. 

The extent to which either of these alternative strategies get used, and what 
consequences they have, are matters that I am in no position at all to say. I 
have little bits of data like this, and not anything much larger than that. But 
in very different materials one finds the same preoccupation. So it may be that 
the conflict that’s set up here is not really terribly atypical - if that’s at all 

Lecture 11 

Espousing a rule ; Exemplary 


Considering that statement, “If you go out hotrodding you’re bound to get 
caught and you’re bound to get shafted,” there are some things about it that 
we ought to note. First of all, the maxim that quite obviously follows from 
it - that you ought not to do the thing - is itself not asserted. So that what 
we get is the parts of a syllogism device as the basis for finding that maxim 
as appropriate. The character of that maxim - that you ought not to do it - 
is something that we could consider by reference to the fact that it isn’t 
asserted, as well as by reference to some other features of the way it’s used. 

One thing is, that with a maxim like that one, even supposing that 
someone proposes it overtly, there are some alternatives as to what’s to be 
made of it. For example, it could be treated as a more or less purely affiliative 
statement, i.e., something prefaced by ‘I believe, and I am a member of those 
who hold the belief that this is or is not a proper thing to do.’ Where, then, 
an answer might be an assertion that those who hold that belief are members 
of some group that we don’t affiliate to. 

It is to be noted that in the variety of maxims that are implied by Joe’s two 
positional statements, they are not explicitly proposed by reference to the fact 
that he thinks they ought to be followed because they are, e.g., good rules, or 
whatever else. So the advice he can be offering is the type which both the 
offerer of, and the one who accepts, can offer or accept without any 
commitment to those persons who have built them, those persons who 
enforce them, etc. They’re only ‘good advice’ to anyone who lives in a world 
which is at least partially governed by the fact that if you do ‘A,’ ‘B’ is liable 
to happen. That is to say, Joe isn’t saying that you ought not to do it because 
it’s wrong - where, if he said that, then they could simply claim that he was 
a traitor among them, or perfectly clearly an adult in child’s guise, or 
something like that. While the technique he’s using does not foreclose such a 
possibility, it is very generally used. That is, the claim that some rule which 
one does not claim is correct or good or whatever, ought to be followed by 
reference to whatever it is are taken to be the interests of the persons being so 
advised. Presumably it’s a major technique for social workers dealing with 
street gangs and that sort of personnel. 

In fact, the character of the warning or threat which is asserted via such a 
technique, is the basis for the kind of phenomenon most technically explored 
in the works of Thomas Schelling. These are a series of war-game type things, 
one very classic technique, which has a weak form and then a much stronger 



Part II 

form, goes like this: You can listen on the radio to the announcements of 
government leaders, and they will often make statements of the following 
form: “If the US keeps on its present course,” say the Russians, “then it’s very 
likely that war is inevitable.” Where they’re making what looks like a mere 
prediction, the follow-up of which they have nothing much to do with. 
They’re not proposing you ought not to do it, or that they think what you’re 
doing is bad, or that they will have anything to do with those consequences. 
It’s a natural fact of life. And for that kind of thing to get strengthened, the 
classic military procedure is, I suppose, the kind of thing that was in the 
movie, Dr. Strangelove , but can be found in places like Schelling, and that is, 
to set matters up so that the consequences you propose naturally to happen, 
will happen as an inevitable consequence of some event. So that if, for 
example, a bomb explodes in a certain place, it will automatically and beyond 
the control of anybody involved, set off a reprisal which has been electronically 
geared to this possibility. Governments tend to do that sort of thing, and the 
US, for example, uses the phenomenon of public opinion in that fashion, when 
they propose that it’s necessary, given public opinion, that this or that military 
strategy be enforced. It does not matter, apparently, whether the consequences 
function electronically, or whether it is simply a matter of the fact that certain 
events will have their emotional consequences for a public - allegedly or in 
fact. And it’s not a matter, either, of whether you - whoever it is who’s 
proposing this thing - are not the one who mobilizes that, and in fact can 
ignore it. 

So statements of this form seem to set up problems, the solution of which 
consists of making them more and more descriptive in fact. That is, because 
these are problems, if you can satisfy their conditions, then you’ve got a 
powerful maneuver; so you arrange things so as to satisfy their conditions. If 
you could have a statement to the effect that if the Soviets move their troops 
in the following way, then, without respect to what you can do, the following 
will happen, and they can see that it’s so, then of course it pays to arrange 
things so that that maxim now holds. 

And relatedly, it is important that if one is attempting to get a set of persons 
to modify the behavior of some others who would stand in opposition - or who 
certainly stand in no commitment - to the set of rules you hold, then you want 
to be able to permit those who are going to espouse those rules to come on as 
perfectly well understanding the circumstances of those they’re talking to, 
perhaps even affiliating with them, but in any event, not at all committed to 
the correctness or the moral rightness of the positions they’re espousing. 

In, I believe it was the 6th century, the Catholic missionaries who went to 
Catholicize England, wrote to the Pope saying, “These people already have a 
religion, and it has all sorts of features to it; for example, they have places 
where they worship, etc. What do we do about it, do we destroy them? After 
all, they are heathenish places, and how in the world can they be accepted or 
acknowledged by us.” To which the answer came back: “Leave them. In fact, 
encourage them. What you want to do is fit our terminology and our ways to 
whatever given ways these people have. We ll be around a lot longer than that 

Lecture 1 1 


one will, and over time they will have forgotten what the other religion was, 
and we will nevertheless have maintained whatever power it has over them.” 

What the constraints are on human nature, is something that we want to 
discover via an investigation of how it is that they in fact do whatever they do; 
where what we’re trying to do is to see what is it and how is it that some ways 
they have of doing something works, and what its import might be. Now 
here we start out with a routine argument of a fairly trivial sort: If you do X, 
Y is bound to happen. There are reasons why it seems to be a quite powerful 
one. We can get lots of examples of it. We can find that it’s used by persons 
who hold to rules without regard to whether they’re efficacious or not, and 
they offer the efficacious formulation to those who don’t hold the rule at all. 
For one, that permits the ones it’s been offered to, to accept it and to 
proselytize it without themselves changing their affiliation. If each person who 
accepted the rule had to espouse it for its moral correctness, then each time 
someone accepted the rule, he might now be expelled from the group he was 
in, in the first place. Suppose you convinced a kid that he ought not to hotrod 
because it’s bad. Each kid who was so convinced, might no longer have any 
access to the hotrodders, who would treat him as a turncoat. What are the 
circumstances under which somebody espouses a view that’s not held by a set 
of members, and can still talk to them? Just consider the question that’s in the 
air about these guys coming back from North Vietnam now. What are the 
circumstances under which they could possibly be in a position to say the war 
ought to be ended? 

Now, one of the other things I want to examine for this is, it is perfectly 
well possible that no matter how you happen to espouse a rule - that is, for 
its moral correctness or its efficaciousness - you’re liable to face the 
circumstances of being called a member of the group that espouses that thing, 
and when you’re called that, the rule is thereby treated as irrelevant. The way 
that happens has to do with the kind of things that, in their queer way, 
philosophers have been much concerned with, at least early in the century. 
And that is, as they put it, if you assert some moral rule, are you doing 
anything more than asserting your affiliation? 

Pretty much all that is interesting about that is, that it is indeed the case 
that the way members account for and, often, explain away some asserted rule 
is by aserting that category for which it is the case that its members hold that 
rule. They can do that effectively when they propose to be members of 
something that stands in alternation to it. And it is along such lines that how 
it happens that arguments end up in name-calling would be laid out. Given 
a position, a category is assigned in terms of which that position would have 
been generated - where there are alternative categories. A situation is opened, 
then, in which those members who have not declared themselves are now 
warned that the positions they take will involve them in affiliating to one or 
another of the categories that are now present. For each claim that Joe makes, 
then each cateogry that’s offered as an account for those claims is being 
proposed to Bob as, do you want to be called one of those? If so, take a stand 
with this guy. If not, don’t say anything, or oppose him. 


Part II 

There are two different ways Joe goes about making his position. One is 
that statement that has as its obvious inference the maxim ‘don’t go 
hotrodding.’ The other is through this hypothetical singular occurrence, 
which it seems he’s proposing to be fully generalized. ‘If I . . . etc.,’ where the 
‘I’ is now merely a person in this thing, where what he says holds for 
whomsoever. And that generalizability is of course attacked in the return to 
it which says it’s not generalizable i.e., ‘Not many people . . . etc.’ 

We could call that second statement of Joe’s either an exemplary 
occurrence or a hypothetical exemplary occurrence, and wonder whether it 
matters that it’s hypothetical or not. And indeed, it needn’t even be clear 
whether it is hypothetical or not, because he can use the ‘if form for an actual 
occurrence, simply to show he’s going to make an argument from it; where 
you use the ‘if form in order to make it stronger. That is, the sheer fact it 
occurred may have no consequences. If you just say, “On a given day the 
following happened to me,” the hearer can say, “So what?” If, however, you 
say, “If I do the following ...” then you’re making of it a general argument. 

Now the phenomenon of ‘exemplary occurrence’ is fundamental for the 
following kinds of reasons. It seems to be the case that a large amount of the 
knowledge that Members hold about how the society works are, as I earlier 
put it, ‘protected against induction.’ And what that means is that it isn’t 
automatically modified if events occur which it doesn’t characterize. The 
question is, does that mean that there’s no way that it can get modified, and 
if not, how does it get modified, where the question of how it gets modified 
can be quite important. 

There could be all kinds of models of how it might get modified. One 
might be, you store up a bunch of exceptions, saving each one, remembering 
them, and when you get a whole bunch of them, you’re forced now to say that 
what you supposed were so is not so. For example, the first Negro you meet 
turns out not to be what you know about Negroes. You store that up as 
information. When you’ve met ten, and they all turn out to be not what 
you’ve been told, you say, “Well I guess that information is incorrect.” 

Now, things are not done in that fashion. Let me give an account of a 
particular situation, so as to show what the consequences are, of their not 
being done in that fashion. One time I was engaged in trying to see how it is 
that coroners make decisions that somebody did or did not kill themselves. 
Those things are handled case by case, one at a time. A case comes in - let’s 
suppose it involves a body being found with a pistol wound in some particular 
area of the head. The area matters, because the coroner can determine that if 
the wound is in a certain area it could be an accident e.g., the person could 
have been doing such a thing as checking whether there were bullets in the 
gun, etc., whereas if the wound is in other places, then the person must have 
had the intention of killing themselves in that way; that is, they wouldn’t have 
otherwise been holding the gun in such a position as to lead to a wound in that 
area of the head. They go through the cases, and each one may get decided 
via one of these rules. 

If you said to the guy who was making those decisions, “It’s a probabalistic 

Lecture 11 


argument; probably in five cases out of a thousand, it might well be that 
somebody accidentally had the gun in that position,” he’d agree. And now 
there could be a variety of ways of considering that five-cases-in-a-thousand 
phenomenon. One would be to take a thousand cases of this sort, save them 
up, then make your decision on all of them together, the five queerest ones 
will be called accidents, the other 995, suicide. That’s not what’s done. 
What’s done is, each case comes up. For each case there’s only a five-in-a- 
thousand chance that it could be not-suicide, so each one is decided to be 
suicide. Furthermore, when a problematic case comes up, what’s said about 
it is, “There’s a very small chance that it could be an accident; it’s not very 
likely, and furthermore we’ve always found that they are suicides.” And the 
fact that, that they’ve ‘always found’ consists of their procedure for deciding 
that, does not seem to make any given determination problematic. That 
they’ve ‘always found’ is used to, in fact, urge that the next case is like those 
that we’ve found. So you don’t get the adding up of the troublesome ones, 
about which can be said, “Well, there are always some that are sort of 
troublesome but they have always turned out to be suicide.” I take that 
because it’s a rather stark example, but consider how people deal with 
excuses, and a variety of other such phenomena; they do it in the same way. 

Now then, the problem is how do these shifts take place in that knowledge 
that gets used on each case, and for which the fact of exceptions occurring 
turns out not to be relevant in each case; even though they will be fully 
recognized as sometimes occurring. To extend the point a bit, consider for 
example, those things which I’m sure have happened to you, where you know 
that such things happen, you’ve heard that they happen, people tell you that 
they happen - except that when they happen to you, you also know that you 
could never tell anybody that they, in fact, happened to you. That is, they 
happen. But they happen to nobody, on no given occasion. And you realize 
you can’t tell anybody about it, except, maybe, to say that it happened to 
somebody else. 

Now the phenomenon of an exemplary occurrence is apparently a kind of 
way that a large shift can take place. And in the case of suicides, just to 
continue that along, things could happen as follows: Until now, deaths at 
railroad crossings, when a car gets hit by a train, are presumed to be accidents. 
Then somebody says, “We’ve had a whole bunch of these recently, they’ve all 
been decided as usual, but this one looks interesting.” They examine that one, 
and it turns out that it’s extremely unlikely that it could have been an 
accident; the train could perfectly well be seen, and, further, the road was so 
situated that the car could not have been seen from the train. Suppose also 
that they do what has been done by people like those in the Harvard School 
of Public Health; say, take the car apart afterwards. They find that there 
weren’t any brakes at all; the driver had taken the brakes out so as to prevent, 
even if they’d wanted to change their ways at the last moment, the car 
possibly stopping. Then, through such a single case, the rule might get 
changed. Now it’s the presumption that if there’s a railroad crossing accident 
of a fairly general sort, it is suicide. 


Part II 

That matter may seem to be not terribly consequential in general, except 
that there can be a whole set of circumstances where a rather general and 
important moral rule of one sort or another gets handled in the same fashion. 
The import of certain kinds of crimes, when they get publicized to some 
extent, is that persons now begin to talk of the outcome of that trial 
implicating how it is that the legal system works. So people said about the 
Chessman case during its course, that if he didn’t go to the gas chamber, then 
everybody would lose respect for our legal system, we would be giving free 
play to deviants, murderers, whatever else, and that everybody would take it 
that they could no longer hold the rule that the legal system by-and-large 
works correctly. 

That means, if exemplary occurrences can be formed up out of a single 
occurrence, and if they can have the consequence of changing some people’s 
knowledge, where furthermore the new piece of knowledge will have exactly 
the same form as the old - that is, it also will be protected from induction, 
then matters are extremely tender. You don’t get a step-by-step modification 
of something. It’s frozen, it shifts, and it’s frozen again. Then the focus on 
single occurrences is not that they are an exception, or that they slightly 
modify what you know. It can be an attempt to shift whatever anybody has 
known previously. 

Lecture 12 

‘Tearing down;’ Non-translatable 


I’m going to finish up our consideration of this segment by dealing with some 
of its more overall features. We could say sort of roughly that the segment 
itself has a kind of introduction, which is that one sentence that they all put 
together, then there’s an argument for a bit, then what may be a transitional 
segment which is that three-utterance interchange, “Face it you’re a poor little 
rich kid,” “Yes Mommy,” “Face the music,” and following that, what they 
eventually call this ‘tearing down.’ 

Now ‘tearing down’ is a procedure that can be done on a large variety of 
occasions and in a large variety of settings, and one of the things we want to 
ask about it is, Does it have any special place in the therapy situation, for one, 
and at this point in the therapy situation — whatever point this may be. The 
way I’m going to approach that is to ask whether that procedure which they 
call ‘tearing down’ has any interesting relationship to what it is that at least 
the patient in a therapy situation takes to be going on, at least some of the 
time. One reason we could have for asking that sort of question might be, we 
might feel as I feel that this segment is a kind of sample presented to the 
‘company’ or new candidate, as to what it is that happens here. And we want 
to see in what way it might be considered to be a sample. 

Here is a piece of a conversation which takes place between a staff member 
of a psychiatric agency and a woman who has, as she takes it, very serious 
troubles, but who is very hesitant to go back into therapy. There’s an attempt 
to convince her to do that, but she offers a characterization of therapy as she 
experienced it before. She says, 

A: I stopped at a very bad time, I know I did. 

B: Uh huh 

A: But the job situation got moving and I I stopped right when I was 

beginning— well I would have been better if I had no therapy than to 
stop when I stopped. 

B\ Yeah, sounds like it. 

A: Cause 1 stopped right when I was looking at the whole gory mess. 

B : Uh huh 

A: Before it started, you know, before I started to clean it up a little. 

B : Uh huh 

A: Was like a surgeon getting down to the disease, you know, and all of 

a sudden he opens you up and there’s disease. Wow. 



Part II 

B : Uh huh 

A: He knows what it is and quits there, (laughs) 

B: Yeah, I suppose that’s it. 

A: (laughs) 

B: That’s a good example. 

She offers this image, then, of being cut open. And that image of first being 
torn apart as a first part of a therapy procedure, is one that, at least patients 
hold often, anyway. What, in part, it consists of is, apparently, taking 
varieties of things that the patient says about whatever, and either explicitly 
formulating for them, or in some way leading them to see, that some set of 
terms characterize their circumstances, where the terms can be more or less 
technical in one or another way. Like, in the case I just quoted, right at the 
beginning the doctor says, “Tell me about your problem” and she says, “I’m 
ashamed to. I don’t want you to tell me I’m infantile, because I know that.” 

That obviously can stand in some fairly nice relationship to what transpires 
in this segment, and her characterization of what happened to her might stand 
in some very nice relationship to that ‘tearing down’ procedure. And many 
analysts of initiation procedures in general, and the early stages of things like 
being hospitalized in mental hospitals, going through therapy, etc., talk of 
them as degredation, mortification, stripping of the personality, and the like; 
any of which might well be seen as the same thing named by the phrase 
‘tearing down.’ But there’s a really deep problem involved. I don’t know the 
best way of presenting the problem involved in sentences that have the form, 
“Don’t tear him down” or “I don’t want you calling me infantile,” but I’ll 
try a couple of ways. They may not seem initially to be just in point, but hold 
on to it for a bit. 

At the beginning of a book called Excess and Restraint by an Australian 
anthropologist, Ronald Bemdt, we find him doing what anthropologists 
occasionally do, and that is, to formulate what it is that he takes it he was seen 
as by the natives when he arrived at this place that he went to work in. This 
is from page ix of the preface: 

We were viewed as returning spirits of the dead who had forgotten the 

tongue of our fathers and wanted to relearn it. 

Grammatically, that’s perfectly good English. I want to make the case that it’s 
an asemantic statement. That is, it is not meaningful in English, though it 
appears to be. 

The first part, “We were viewed as,” is fine, and there are lots of things 
that can be stuck into that for which this first part is a frame. And that frame 
to some extent controls what can be put into the second part, in the sense that 
there are lots of things that could be used as second parts of sentences, that 
don’t fit here, syntactically or otherwise - for example, that isn’t the kind of 
thing you say about humans. 

I suppose it has to be clear that the second part, from “returning spirits” 

Lecture 12 


on, is some kind of translation from another language into English, and, in 
part, that it is a translation is a considerable part of its warrant for being said. 
Such that, if Berndt stood up at an anthropological convention and said, “I, 
Ronald Berndt, am a returning spirit from such-and-such,” that would be 
good grounds for them to think he’s insane. That is to say, if this were 
English, it could not be a potentially correct description of what he might be. 
But there are apparently ways that, if you take words one by one and in some 
combinations, some combination of them can be produced to look something 
like a Members’ category. Those words could somehow go together, and be 
recognized as ‘a something in some world.’ 

We have some way of handling such a statement because you can say, 
“Well, the initial thing is that he’s a stranger, and that’s some way of 
characterizing strangers that involves somehow making him some sort of 
member, and that must be one category that these people either had or 
devised for handling such persons.” But we would take it that, except as a 
matter of syntax, none of the controls on correct use of that clause ‘returning 
spirits . . . etc. ’ are found in ordinary English. We would not know when it’s 
been said correctly about someone, what are the consequences of it being said, 
what you can infer from its being employed, etc., etc. It would be a sentence, 
then, if it had any distribution in English, which would be reporting what it 
was that was taking place for these other people. And it isn’t a proposed 
introduction into English, I take it, of some category that we’ve lacked and 
now are enriched by having. So, for example, until the term ‘schizophrenia’ 
was invented, it could be said that we lacked that category, and that given it, 
a body of experience, a set of observations, etc., could now be ordered, and 
‘schizophrenia’ now stand as their name - and be used, furthermore, as other 
disease names might get used. But what kind of experience this term 
‘returning spirits . . . etc.’ orders, and what class it would be a part of is, at 
least to me, quite obscure. And yet there’s a sense in which we do understand, 
by some mapping, what sort of thing he is talking about. And that, of course, 
is an extremely interesting kind of fact. 

The sentence, then, is an odd one, and the introduced category is not, as we 
might put it, generative. That is to say, it’s not like a Reader’s Digest word 
which, now having learned it, there are lots of places where you can use it. But 
there may be somehow in there, an exploitation of English syntax in making 
the translation and building a sentence like that. And that is, in part, I 
suppose, why anthropologists will frequently not do this, but will say, ‘We 
were viewed as . . .’ and give you some phrase in the native language, and 
then say explicitly, “What we have here is a gloss of something that might 
mean the following sorts of things in English.” But they still use the native 
term, because then you can’t use the term in English; that is it remains a term 
from that other language. 

I introduce this sort of consideration since I want eventually to make the 
case that one of the core problems for psychoanalysis and other psychother- 
apies, and the social sciences in general, concerns the relationship between the 
categories that those disciplines set up and the categories that members of the 


Part II 

society otherwise use. The problem is that somehow members take it that 
such categories - ‘manic depressive,’ etc. - are additions to a list of categories 
that exist already, and can be used in just the same fashion that the old ones 
are usable. They may be better, but they do not otherwise modify the 
structure of the class, of which people come to be seen as members. However, 
the professional constructing these new categories may take it that one major 
task he has is to somehow build them so they are unusable in the way that the 
categories he sees them as replacing were usable. That is, the professionals put 
it as a programmatic task that they would like to have it that the statement 
“You’re a manic depressive,” for example, would be nonsense in ordinary 
English, i.e., unless said by a therapist. 

When Freud set out to build a ‘scientific psychology’ as he put it - and that 
program of his, to build a scientific psychology, was among the last things he 
ever got translated into English - one of the tasks he felt he faced in the first 
instance, was to deal with the fact that everybody considered themselves to be 
an authority in psychology. He was not locating a domain which persons 
could take it from the beginning they had no idea what it is you might say 
about it. But it wasn’t, of course, only the case that they had lots of views, 
which they took to be well warranted, on psychological matters, but those 
views, and the categories they used, were not morally neutral. That is to say, 
in part, that any time one of them was asserted as being so about somebody, 
it was also the case that some assessment of their status was being made - as 
a good citizen, or whatever else. 

Now, in many views of what could be called the philosophy of science, it 
is taken that technical findings stand in some position of replacement to lay 
findings. One can often have a picture of the state of lay knowledge which 
stands in a strong relationship to what scientific knowledge is supposed to 
look like. That is to say, that it’s a set of items about any or every given topic, 
where what happens is, when a new item is introduced, it can operate to 
expunge an old item. That goes for correct scientific procedure, and it has 
been taken to go for what is called lay knowledge, as well - with some delay 
perhaps, but essentially the same kind of thing can properly take place. 

That’s a kind of tricky argument if you’re dealing with the structure of lay 
knowledge, although it can be perfectly well satisfactory if you’re dealing with 
the structure of scientific knowledge, in that you can set up by fiat what the 
structure of scientific knowledge is, and you might get a list of items, and you 
might, then, at any time, remove an item as a new one comes in that seems 
inconsistent with it. However, what the structure of lay knowledge is, would 
obviously be a problem for some discipline. It’s not in the first instance 
known. It clearly has been a subject of some consideration since the Greeks, 
anyway. And to propose that it’s a list of items is to make a rather daring 
assertion. But it’s one that has been made. 

As a further problem, there is the fact that it seems if an item is now 
introduced to the scientific corpus, then the controls on its use can presumably 
be very well laid out, and whether it is correctly used can be something 
insisted on and enforced. And furthermore, it can be presumed that if it’s 

Lecture 12 


incorrectly used, you will in fact have hell to pay; that is to say, consequences 
will just show up if you misuse some item. Now it’s not clear, in the first 
instance, what kinds of uses these items do have when we re talking about lay 
knowledge. It’s by no means clear, either, that they’re only the same uses, or 
that ‘error’ - if we can formulate that - has the same short-run or long-run 
chance of showing up. 

For considerations like the ones I’ve offered, there is some considerable 
concern about who would have the right, at all, to use new scientific 
knowledge, or techniques, or whatever. Because, for one, if ir became 
something that anybody could use, then God only knows what might be done 
with it. And the simplest way for it to become something that anybody might 
use, is for it somehow to be seen as a member of the corpus of correct 
knowledge, undifferentiated as to scientific or lay. Where it’s the fact that 
scientists say so that permits anybody now to assert and use it - where, 
however, they need make no special change in how they use any category, to 
use the new one. 

One thing Freud was further concerned about was to somehow prevent 
persons from using his categories in just the fashion they used the ones they 
had before. And there’s very good reason for that, which is that if, say, ‘manic 
depressive’ was a replacement for some lay term like ‘cranky,’ then whatever 
assessments that were made about somebody said to be ‘cranky’ could be 
made about a person said to be 'manic depressive,’ and that someone who was 
said to be manic depressive might hear it as a kind of attack. 

Now, I take it that, probably genetically, in the first stage of therapy, 
anyway, that’s what’s seen as going on. And the term ‘infantile,’ for example, 
is treated as being either the same as the lay term ‘infantile,’ when someone 
says to an apparent adult, “You’re behaving like an infant,” or as some sort 
of technical term which is nonetheless usable as the other term is. And that is 
something that, for example, let’s say a field like astronomy has avoided, and 
nobody gets very horrified to hear about the generative stars. But it’s not been 
done in such fields as psychiatry. And the fact that it’s not been done has been 
turned into a tremendous attack on psychiatry, by somebody like Thomas 
Szasz, who will invoke the fact that the categories are used evaluationally 
elsewhere, to propose that the phenomenon is that of making an evaluation, 
and that their isn’t anything but that sort of knowledge being used. 

That is, I suppose, a strong indication of the failure to cut off the 
incorporation of new information into lay hands. And it would be in that sort 
of way that one can see how the so-called ‘same procedure’ can be seen as one 
that ‘you call X and we call Y.’ You call it ‘diagnosis’ and we call it ‘tearing 
down.’ And one can see how, then, these patients could be characterized as 
giving their version of what it is that’s going to happen in these sessions. It 
remains the case that if the discipline retains the notion of items of knowledge 
as the things it is basically dealing with, then it’s going to have to face the 
problem that the items get taken and used. One thing that can be done, of 
course, is to attempt to place anything that could be called such an item into 
such a structure that if you pull the item out, it’s senseless. 

Lecture 13 
‘Everyone has to lie ’ 

The transcript of lecture 13 is incomplete and full of gaps. What there is of it 
deals with the phenomenon that Sacks called ‘ Everyone has to lie.’ It became a 
paper of the same name, and can be found in M. Sanches and B. G. Blount (eds), 
Sociocultural Dimensions of Language Use (New York; Academic Press, 
1973), 37—80. For the materials in lecture form, see Spring 1967 , lectures 8 
and 9. 


Lecture 14 
The Navy pilot 

The transcript of lecture 14 is full of gaps. In its place are Sacks' notes on the 
data he considered in lecture 14, a New York Times article entitled ‘A navy 
pilot calls Vietnam duty peak of career. ’ The materials are taken from Sacks’ 
Research Notes, vol. 2a (1964—5), 170—99. 

MCD materials 1 

New York Times Sat. May 29, 1965 
From a long article: 

A Navy Pilot Calls Vietnam Duty Peak of Career 

by Seth S. King 

Special to the New York Times 

ABOARD THE CORAL SEA, in the South China Sea, May 27 

Commander Jack H. Harris, leader of the Attack Squadron 155 aboard 
the Coral Sea was explaining how a carrier pilot feels about Vietnam: “I 
certainly don’t like getting shot at. But this is the top of my flying career 
and it’s important I should know about it. I really feel I’m fortunate to 
get the opportunity after 2 1 years in the Navy of combat experience. I 
need it to be a real professional.” 

How did he feel about knowing that even with all the care he took in 
aiming only at military targets someone was probably being killed by 
his bombs? 

‘‘I certainly don’t like the idea that I might be killing anybody,” he 
replied. "But I don’t lose any sleep over it. You have to be impersonal 
in this business. Over North Vietnam I condition myself to think that 
I’m a military man being shot at by another military man like myself.” 

1 First, we may see one crucial matter here; and that is that he takes it 
that there are alternative ways that he and those he is dealing with (bombing, 
being fired at by), may be categorically formulated. 

1 ‘MCD’ stands for Membership Categorization Device. 



Part II 

2 Second, that the choice of the categorizations is relevant to the 
assessments he might make of his activities; now this is so in a strong sense. 
To wit; that he sees that the formulations have alternative imports, and it is 
not just the case that the formulations have alternative imports, and that 
further by some work he is able to choose that formulation which has an 
import that is consistent with what it is that he is doing, and that permits him 
to do it without important moral consequences (he is able to continue 
‘without losing sleep’). This situation of alternatives is then one to be marked 
as one where he takes it that there is not some exclusively appropriate or 
required choice of devices. 

3 It may be much noted that the choice of the device is not just to be 
made by reference to what they may make of his actions under some scheme 
like reference to their consequences or the like, but by reference to how, given 
the use of the consistency rule to formulate his alter egos, how it provides for 
the formulation of his actions. The availability then of making his categori- 
zation decision in such a way that it routinely provides for a categorization of 
his opponents, and by virtue of the mutual categorization then an assessment 
of either’s actions is a crucial matter. 

4 Notice how it is that categorization can be held in some abeyance or the 
use of particular devices not be foreclosed while making reference to 
populations, i.e., the use of ‘anybody’ and ‘someone.’ 

5 Notice also how the characteristics of the required impersonality are 
delivered, via the use of ‘you’ in a usage we have been considering. 

6 There are several matters re ‘you have to be;’ is it the case that if you 
have to be then you can be; or is it the case that, that you have to be provides 
the grounds for being. 

7 Now the import of the categorization of the others as well as himself, 
and of the particular device chosen, is in part that an ethic follows, where: 
what is proper for him is proper for them, or, equally, what is proper for them 
is proper for him; and also, that the device chosen is not merely one which 
assigns a category to he and to them, but also it offers a category to them, 
which they may also use in just the way he uses them, i.e., that with it they 
may formulate what is happening just in the terms he does, and with the 
same consequences. If they accept that offer then at least there are no 
complaints to be offered on their part about the error of his ways, except if 
he happens to violate the norms that, given the device used, are operative; 
that is, the device used does not provide that there are no rules for the actions 
of either. 

It is of course to be noted about the given device that it delocalizes the 
affiliation of the personnel; what is relevant in the first instance is not that we 
are the military of the US and they of Vietnam, but that we are both military; 
it is that we are differentially affiliated which provides for who we shoot at; 
otherwise we are colleagues. 

8 One of the things to be noted about ‘you have to be impersonal in this 
business’ is that its indexical features, particularly the ‘this,’ have some special 
relevance. The statement is perhaps only incidentally correct for the military; 

Lecture 14 


it is correct for other businesses as well. And that feature of it may be quite 
crucial. For, given the fact that it is true for other businesses as well, that the 
consequences of impersonality differ may be rather irrelevant; so, businessmen 
may be impersonal in different ways, or psychiatrists or priests; but if the 
statement is true for them, and they are in a business for which the sentence 
is true, they may not be in a position to without recourse to their own 
circumstances importantly claim that he is doing wrong; for he can answer 
that the same holds for them with the difference being that their impersonality 
doesn’t directly concern killing. The property of prepared, or general purpose 
indexical statements may be a rather crucial one; cf. ‘That’s the problem with 
society’ and others like it. 

And we know that persons who are attacked for the peculiar defects of 
their activities are often seeking to formulate matters such that they can 
claim that the same sort of defect, if not the details of it, hold elsewhere, 
for the attacker, and perhaps for matters quite generally. This use of a 
general purpose indexical statement may be very helpful in such an 
attempt. And note of course that if one can make such a claim, that it does 
seem that one prevents the attack from having its intended power; there is 
something akin to an unclean hands argument. The unclean hands argu- 
ment is further made powerful by then attaching as a further class beyond 
its scope the ‘ivory tower’ or ‘bleeding hearts’ group as further persons who 
are in no position to talk, because they’ve never met a payroll. One needn’t 
then have a single argument to handle every locus of attack; but several can 
be combined; where the first part gathers a bunch together, and then on 
behalf of all, one uses the second part (where one proposes that one is 
defending those who are talking oneself from the first group when one 
attacks the second.) 

It may then be that the use of ‘this’ with its possible extension being greater 
than ‘military’ serves to cut off attacks by anyone who is an incumbent in 
some business for which the statement is also correct. The use of ‘you’ in this 
statement has been examined in the lectures on the first page of the initiation 

9 One can notice that the relevance of a device with which a consistency 
rule usage gives: military-military; is of separate importance than ‘you have 
to be impersonal in this business.’ For that remark might be made by a 
professional gunman, and it would not serve as an answer, as an adequate 
account. Or would it? Or is it relevant? 

10 Is there some suggestion such as: If we are both military men, 
shooting at each other as we ought, then not only is it the case that neither has 
any position to complain about the proper military doings of the other, but 
also what we do is in some sense just our business, and having behaved 
properly, no one else is in position to complain either? Alternatively or in 
combination there is the issue that having taken care of part of their side by 
giving them a military man to behave as I do, he being their representative 
involves them sufficiently; they have no recourse because their side is behaving 
as ours is. 


Part II 

11 One wants of course to notice that the formulation ‘this business’ is 
of interest not only for the ‘this’ but for the fact t